SimBot Challenge Rules

Please review the rules before taking part in the challenge.

The Alexa Prize SimBot Challenge (the “Competition”) is a skill competition sponsored by Amazon.com Services LLC (“Sponsor”), in which competing teams of college or university students build Alexa skills using the Alexa Skills Kit APIs (“SimBots”) for use in the United States, that allow Alexa multimodal device users to maneuver a simulated robot to interact with an environment and complete challenges via Amazon Alexa.

Selected teams (as outlined below) will be chosen to compete in this Challenge, with final results announced in April 2023. See below for the complete Competition details.

By applying to or participating in the Competition, you agree to these Official Rules. Please read them carefully.

COMPETITION CALENDAR: The Competition starts on October 4, 2021 and ends in June 2023 (the "Competition Period"). The Competition phases are noted in the calendar below. The dates are approximate and are subject to change at Sponsor’s discretion.

Phases

Starts

Ends

Phase 1: Participant application period

October 4, 2021

October 31, 2021

Phase 2: Sponsor application review period

November 1, 2021

November 9, 2021

Phase 3: Initial participant notification period/onboarding

November 10, 2021

December 14, 2021

Phase 4: Benchmark challenge

January 3, 2022

April 29, 2022

Phase 5: Limited second participant application period

May 9, 2022

May 23, 2022

Phase 6: Second sponsor application review period

May 25, 2022

June 3, 2022

Phase 7: Second participant notification period/onboarding

June 6, 2022

June 28, 2022

SimBot boot camp

August 8, 2022

August 11, 2022

Phase 8: Initial skills development period

August 12, 2022

October 16, 2022

Phase 9: Skill certification period

October 17, 2022

October 30, 2022

Phase 10: Internal Amazon beta period

October 31, 2022

November 29, 2022

Phase 11: Initial feedback period

November 30, 2022

February 1, 2023

Phase 12: Semifinals interaction period

February 2, 2023

March 22, 2023

Phase 13: Finals interaction period

March 23, 2023

April 28, 2023

Winners announced

June 2023

ELIGIBILITY: To be eligible to participate in the Competition, you must: (1) be enrolled as a full-time undergraduate or post-graduate student at an accredited college or university (other than colleges or universities located in any of the Restricted Jurisdictions defined below) (“Universities”) and remain a full-time student in good standing at such University while participating in the Competition; (2) be at or above the age of majority in your country, state, province or jurisdiction of residence at the time of entry; and (3) not be a person or entity under U.S. export controls or sanctions. Competition is void in Cuba, Iran, Syria, North Korea, Sudan, the region of Crimea, and where prohibited by law (“Restricted Jurisdiction”). People who, during the Competition Period, are directors, officers, employees, interns, and contractors (“Personnel”) of Sponsor, its parents, subsidiaries, affiliates, and their respective advertising, promotion and public relations agencies, representatives, and agents (collectively, “Competition Entities”), immediate families members of such Personnel (parents, siblings, children, spouses, and life partners of each) and members of the households of such Personnel (whether related or not) are ineligible to participate in this Competition. Sponsor reserves the right to verify eligibility and to adjudicate on any dispute at any time.

Entrants (“Entrants”) must enter as part of an “Entrant Team” consisting of one or more students from a single University. An Entrant is only permitted to be part of one Entrant Team. Any Entrant that is part of more than one Entrant Team may be disqualified and his/her corresponding Entrant Teams may be disqualified at the sole discretion of Sponsor. Only one member of the Entrant Team may submit an application on behalf of an Entrant Team but all those listed on the application are Entrants. All prize money payable to Entrant Teams will be split evenly among the Entrants of a winning Entrant Team (based on the pre-tax amount of prize money), as identified in the application, as updated as permitted by these Official Rules, and who maintain full-time student status in good standing at their University during the entire Competition Period.

Each team must select a faculty advisor, who, with respect to that team, will act as the official representative for the Entrant Team’s University (the “Faculty Advisor”). Each Entrant Team must have its own Faculty Advisor; Faculty Advisors may not represent multiple teams. Faculty Advisors are not members of the Entrant Team and will not receive any portion of any prize. Faculty Advisors must remain full time employees of the Entrant Team University during the entire Competition Period. During the Competition Period, but no later than February 1, 2023, the Faculty Advisor may request to remove members from the Entrant Team or to add additional members to the Entrant Team. The Faculty Advisor must provide an explanation of the reason for the removal or addition, and any proposed new member must provide documentation requested by Sponsor and agree to comply with these Official Rules. Changes to the Entrant Team are subject to Sponsor’s approval in its sole discretion. If at any point during the Competition Period, the Entrant Team’s Faculty Advisor cannot continue to serve as a Faculty Advisor, the Entrant Team may submit a request to Sponsor to select a new Faculty Advisor. Changes to the Entrant Team’s Faculty Advisor are subject to Sponsor’s approval in its sole discretion. If an Entrant Team fails to have a Faculty Advisor at any point in time during the Competition, for any reason, the Entrant Team may be disqualified.

Each Entrant must be eligible to participate in this Competition and comply with these Official Rules or the Entrant, and the Entrant Team associated with that Entrant, may be disqualified. This Competition is subject to all applicable federal, state, territorial, provincial, and local laws. Competition is void where prohibited. By participating in the Competition, all Entrants accept and agree to comply with and abide by these Official Rules and the decisions of the Sponsor which will be final and binding including the Sponsor’s right to verify eligibility, to interpret these Official Rules, and to resolve any disputes relating to this Competition at any time. Only Entrants may work on the SimBot, although members of the Alexa team may provide support to the Entrant Team during the Competition Period, and the Faculty Advisor, and other students and faculty members at the Entrant Team’s University may provide support and advice to the Entrant Teams and may co-author the Technical Article (as defined below) or other research papers.

DESCRIPTION OF COMPETITION PHASES:

PHASE 1 “Participant Application Period”: Between October 4, 2021 and October 31, 2021, the leader of each Entrant Team that wishes to enter the Competition may visit TKTKTK (“Competition Site”) where they may submit their entry information via the YouNoodle, Inc. (“YouNoodle”) application portal, including, but not limited to: complete names, contact information and resumes of all Entrant Team members, proof of University enrolled status (e.g., verification of enrollment or an uploaded copy of Entrants’ student IDs), name and contact information of a sponsoring Faculty Advisor from the Entrant Team’s University, and bio for each Entrant Team member.

SimBot Challenge: SimBot Challenge is an interactive, multimodal dialog-driven game challenge in which a virtual task-driven embodied robot in a simulated home or work environment assists a human user in completing real-world tasks by continuously learning, and gaining the ability to perform commonsense reasoning. The robot, an AI-driven virtual agent designed by each Entrant Team, should perceive its visual environment and understand and interact in conversational language with human users. The objective of the game is for the simulated robot to learn and complete increasingly more complex tasks in an efficient manner with help from a human user (player) as needed. Entrant Teams’ robots will need to be able to appropriately perceive, reason, and take actions in the environment to complete all tasks, and interact with the user to clarify ambiguous instructions and obtain additional information when needed. Participating teams are responsible for designing and delivering a multi-modal conversational AI model that is capable of executing and reasoning the above scenarios efficiently to completion.

Each Entrant Team will be required to answer the following questions:

  1. What is your team’s vision for your SimBot?
  2. How will your SimBot help its users?
  3. How do you want users to feel at the end of an interaction with your SimBot?
  4. How would your team measure success in the Competition?
  5. Please share sample interactions you expect your SimBot to achieve by the end of the Competition.
  6. Describe, in detail, your scientific approach, the related system architecture, any innovative claims, technical challenges you will need to overcome, and risks. Plan for operating at scale when you define the architecture. (2 page PDF structured as a technical paper and architecture diagram with 1 additional page for visual or text explanation of key components and data flow).
  7. What is novel about the team’s approach? (This may be completely new approach or a novel synthesis or extension of existing techniques).
  8. How do you think your work will impact the field of AI?
  9. Please attach an overview of the primary roles, subject matter expertise, and focus/specialty of each of your team members. (1 page)
  10. Please provide a summary of the technical work/research (relevant to your proposed architecture), yours or others’, that you will leverage and how.
  11. How will you ensure you create an experience users find engaging?

Entry Applications may be submitted at any time during the Participant Application Period. Only one individual per Entrant Team may submit an Entry Application on behalf of their Entrant Team. Other members of the Entrant Team must accept the invitation to join the team via the YouNoodle application portal. All Entrants must create an account with YouNoodle if they have not done so already in order to submit an Entry Application or accept an invitation to join an Entrant Team and participate in the Competition. Creating and maintaining a YouNoodle account is free of charge. All Entry Applications must be complete when the Participant Application Period closes at 11:59 pm Pacific Time on October 31, 2021. Entry Applications are not complete until all the online prompts and instructions to upload the Entry Application have been properly followed, the Official Rules have been accepted, and all Entrant Team members and the Faculty Advisor have accepted their invitations to join the Entrant Team via the YouNoodle application portal. Entry Applications may not be revised once submitted. Entry Applications will not be returned and become the property of the Sponsor.

Entry Terms: Determination of compliance with entry and other requirements and these Official Rules will be in the sole discretion of the Sponsor.

By entering, Entrants warrant that:

  1. entries are the original work of the Entrant Team or an update to an original work of the Entrant Team;
  2. entries do not infringe or violate the rights of any third party, including but not limited to, copyrights, trademarks or copyrighted material not owned by the Entrant Team, contract and licensing rights, rights of publicity or privacy, moral rights, or any other intellectual property rights; and
  3. entries are not subject to any third party agreements, and that Sponsor will not be required to pay or incur any sums to any person or entity as a result of its exercise of any rights granted under these Official Rules.

Additional Requirements

Entrant Teams must submit their SimBots via Sponsor’s Developer Portal, and in doing so must agree to the Amazon Developer Services Agreement (“Developer Agreement”). SimBots must comply with the terms applicable to “Alexa Skills” in the Developer Agreement, including the content guidelines and all other Program Policies (as defined in the Developer Agreement). SimBots that cause any harm, as determined by Sponsor in its sole discretion, will be disqualified from the Competition. Questionable or controversial designs or content must be cleared with Sponsor prior to being deployed in a SimBot. All SimBots must interact with users using only the English language. SimBots may not request any personal data or personally identifiable information from Alexa users. All Entrant Teams must host their SimBots using Amazon Web Services (“AWS”) Lambda, and in doing so must agree to the terms of the AWS Customer Agreement. However, Entrant Teams may use other frameworks or web services to build their SimBot, and SimBots may make external calls to other services during run-time execution. During all phases of the Competition, Sponsor may in its sole discretion, require all Entrant Teams to provide periodic status updates, reports, or demonstrations of the SimBots. Sponsor may also require all Entrant Teams to comply with additional rules, requirements, or procedures that Sponsor determines, in its sole discretion, are necessary for the administration of the Competition. Sponsor may, in its sole discretion penalize teams for noncompliance with any rules, requirements, or procedures, up to and including disqualification from the Competition.

Sponsor may also, in its sole discretion, require all SimBots to adhere to requirements designed to ensure that SimBots are focused on the goal of advancing the science of conversational artificial intelligence and to ensure a positive experience for Alexa customers using the SimBot. Sponsor may determine any appropriate penalties for noncompliance with those requirements, including temporary suspension of the SimBot, disregarding ratings for interactions that are non-compliant with such policies, or disqualification from the Competition.

Sponsor may provide Entrants selected to participate in the Competition access to software, software development kits, libraries, APIs, documentation, sample code, templates, data sets, and related materials (“Program Materials”) that may be used in connection with the Competition. If an Entrant uses any Program Materials, he or she is subject to and agrees to comply with Sponsor’s Program Materials License Agreement. Program Materials may include APIs, data sets, and other materials that are not public (“Non-Public Materials”). Entrant Teams whose SimBots are made available through Sponsor’s Alexa service (beginning in Phase 10, below) will also receive automatic speech recognition text transcriptions and other data (e.g., confidence scores) relating to Alexa user interactions with Entrant Team’s SimBot (“Interaction Data”). Each Entrant (and each Faculty Advisor) agrees that he or she will not disclose Non-Public Materials or Interaction Data to anyone other than other Entrants of the Entrant Team and the Entrant Team’s Faculty Advisor without Amazon’s prior written consent. Each Entrant (and each Faculty Advisor) also agrees that he or she will use Non-Public Materials and Interaction Data only in connection with the development and operation of (a) a SimBot as part of the Competition and in compliance with these Official Rules and (b) following the completion of the Competition Period, a SimBot or related bot operating as a stand-alone Alexa skill. Entrants (and Faculty Advisors) may not use Non-Public Materials or Interaction Data for any other purpose. Prior to the conclusion of the Competition Period, Entrants may not publish their SimBot as a stand-alone Alexa Skill or use the Non-Public Materials or Interaction Data to operate the SimBot as a stand-alone Alexa Skill or in connection with other Alexa skills. Entrants (and Faculty Advisors) must return or destroy all Non-Public Materials and Interaction Data (in any form and including, without limitation, all summaries, copies and excerpts of the same) promptly following any request from Sponsor, or if the Entrant Team discontinues the development or operation of its SimBot. If any Entrant or Faculty Advisor is disqualified from the Competition, leaves the University, or otherwise terminates their participation in the Competition, that individual must immediately return or destroy all Non-Public Materials and Interaction Data in their possession. Sponsor reserves the right in its sole discretion to impose additional terms and conditions on the use of Program Materials, and to condition access to Program Materials on Entrants’ agreement to those terms and conditions.

PHASE 2 “Sponsor Application Review Period”: From November 1, 2021, to November 9, 2021, all eligible Entry Applications will be reviewed by a panel of judges chosen by the Sponsor. Up to ten Entrant Teams determined by Sponsor according to the criteria described below will be chosen to participate in the Challenge described below. To select the Entrant Teams chosen to advance to the Challenge, Sponsor will review each of the Entry Applications using the following criteria:

  • The potential scientific contribution to the field;
  • The technical merit of the approach;
  • The novelty of the idea; and
  • An assessment of the Entrant Team’s ability to execute against their plan.

PHASE 3 “Initial Participant Notification Period and On-Boarding”: Leaders of Entrant Teams that have submitted Entry Applications and have been selected by Sponsor to participate in the Competition will be notified by email at the email address provided at time of Entry Application between November 10, 2021, and December 14, 2021. Entrant Teams acknowledge that some Entrants and Faculty Advisors may have participated in prior iterations of the Alexa Prize competition (“Prior Year Teams”).

Stipends: Each Entrant Team selected to participate during the Initial Participant Notification Period will be eligible to receive a stipend award of $250,000 U.S. dollars (paid in any number of installments as determined by Sponsor in its sole discretion, subject to Entrant Team’s continued participation in and eligibility for the SimBot Challenge), which will be awarded to the Universities of the selected Entrant Teams in the form of restricted research grants. These stipends will be awarded to the Universities and not the Entrant Teams or any individual Entrant, and will be subject to the University signing and returning any documents required by Sponsor (including IRS forms W-9 and/or W-8) and to the University agreeing in writing that no more than 35% of the research grants may be allocated to administrative fees. Entrant Teams will not be eligible to begin participating in the Competition until all required documentation has been completed by the University. The grants will be awarded to support two full-time students or the equivalent of two full-time students during the Competition and one month of the Faculty Advisor’s salary. Each University will be responsible for allocating and managing the funds within these guidelines and for payment and reporting of any required taxes, withholdings, fees or duties. Sponsor is not responsible for ensuring all or any portion of stipend research grants are distributed to Entrants. Each member of an Entrant Team whose University receives a stipend award, and its Faculty Advisor, will also receive Alexa-enabled devices (up to a maximum of four devices per Entrant Team, including the device provided to the Faculty Advisor), free AWS services to support the development of their SimBot (subject to reasonable limitations set by Sponsor), and support from the Alexa team as determined by Sponsor.

Stipends are non-transferable except as directed by Sponsor. No stipend substitutions or cash redemptions are allowed except as designed by Sponsor. Except where prohibited by law, all federal, state, provincial or other tax liabilities or withholdings are the responsibility of the University and the Sponsor will not be responsible for any tax deductions which may be necessary, except that Sponsor may withhold taxes as required by law. Universities are responsible for any costs and expenses associated with stipend acceptance and use. If an Entrant Team withdraws from the Competition or does not remain compliant with these Official Rules, Sponsor will be relieved of any obligation to pay any remaining portion of the stipend to the Entrant Team’s University. All details of stipend not specified herein shall be determined solely by Sponsor.

PHASE 4 “Benchmark Challenge”: From January 3, 2022 through April 29, 2022, selected Entrant teams will participate in a Benchmark Challenge. During the Benchmark Challenge, participants will build preliminary SimBots (machine learning models that can complete goals given a dialog history and scene images) using a dataset released by Amazon. The Benchmark Challenge will be publicly hosted on evai.ai, and will be subject to and governed by the Benchmark Challenge rules that will be posted on eval.ai no later than January 3, 2022. Because the Benchmark Challenge will be open to the public, teams may participate even if they were not selected for the SimBot Challenge. Following the Benchmark Challenge, Amazon in its sole discretion may select up to four additional teams to participate in the Simbot Challenge, in accordance with the process below.

PHASE 5 “Limited Second Participant Application Period”: From May 9, 2022 through May 23, 2022, teams that successfully participated in the Benchmark Challenge during Phase 4 may apply to be selected to participate in the Simbot Challenge. Entrant Teams previously selected in Phase 1 do not need to re-apply. As in Phase 1, new Entrant Teams must submit their entry information and responses to the entry questions via the YouNoodle, Inc. (“YouNoodle”) application portal, following the process set forth in Phase 1 above. All application procedures, Entry Terms, and Additional Requirements set forth for Phase 1, above, apply in full for Entrant Teams applying during Phase 5. All Entry Applications must be complete when the Participant Application Period closes at 11:59 pm Pacific Time on May 23, 2022.

PHASE 6 “Second Sponsor Application Review Period”: From May 25, 2022, to June 3, 2022, all eligible Entry Applications will be reviewed by a panel of judges chosen by the Sponsor. Up to four Entrant Teams determined by Sponsor may be chosen to participate in the Challenge described below. In addition to the criteria considered during the Sponsor Review period in Phase 2, the judges may consider the Entrant Team’s performance in the Benchmark Challenge when selecting additional participating teams.

PHASE 7 “Second Participant Notification Period and On-boarding”: Leaders of additional Entrant Teams that have submitted Entry Applications and have been selected by Sponsor to participate in the Competition will be notified by email at the email address provided at time of Entry Application between June 6, 2022, and July 15, 2022. Entrant Teams acknowledge that some Entrants and Faculty Advisors may have participated in prior iterations of the Alexa Prize competition (“Prior Year Teams”). Each Entrant Team selected to participate during the Second Participant Notification Period will be eligible to receive a stipend award of $200,000 U.S. dollars (paid in any number of installments as determined by Sponsor in its sole discretion, subject to Entrant Team’s continued standing in the Simbot Challenge), which will be awarded to the Universities of the selected Entrant Teams in the form of restricted research grants. All restrictions on Stipends set forth above for Stipends awarded during Phase 3 will apply in full to Stipends provided to Entrant Teams in Phase 7.

SIMBOT BOOT CAMP All Entrant Teams will be invited to an Alexa Prize Simbot Boot Camp held in August 2022 at Sponsor’s headquarters in Seattle, Washington or alternative location as determined by Sponsor. Sponsor will provide for the cost of the Faculty Advisor and four members from each Entrant Teams’ airline tickets (non-refundable coach class booked through Sponsor at least 14 days in advance) from the city where their University is located to Seattle, Washington, and the provision of hotel rooms in Seattle, Washington, to permit members of each Entrant Team (up to 4 members as determined by Sponsor) and the Faculty Advisor to attend the Boot Camp. Travel expense subsidies may be subject to tax information reporting and withholding to the extent required by law. Attendance at the Boot Camp is not required and will not affect an Entrant Team’s opportunity to win. Sponsor may also choose to hold the summit virtually and in which case Entrant Teams will not receive any payment for the value of travel expenses.

PHASE 8 “Skills Development Period”: From August 12, 2022 through October 16, 2022, selected Entrant Teams will develop their SimBot. Prior Year Teams may utilize and incorporate any concepts, works, inventions, information, designs, programs, or software created by Prior Year Teams, provided that the Entrant Teams have obtained all of the rights, licenses, and permissions in writing from any person who may have helped create the SimBot or that are otherwise necessary for the Entrant Team to distribute the SimBot and to grant the licenses set forth in these Official Rules and in the Developer Agreement. Any Entrant Team may incorporate concepts, works, inventions, information or designs described by any Prior Year Team in its published technical articles. Entrant Teams must submit their SimBot to Sponsor for certification no later than 11:59 pm Pacific Time on October 30, 2022.

PHASE 9 “Skill Certification Period”: From October 17, 2022 through October 30, 2022, Entrant Teams must revise their SimBots in response to feedback from Sponsor to ensure that their skills are certified prior to being made available to end users of the Alexa service. SimBots must pass certification in order to be eligible to participate in a closed beta program for employees of Sponsor (the “Beta Program”).

PHASE 10 “Internal Amazon Beta Period”: Beginning on October 31, 2022, SimBots that have passed certification will be made available via Sponsor’s Beta Program. Each SimBot’s Interaction Ratings (as defined below) collected through the Beta Program will be provided to the Entrant Team. Entrant Teams may continue to improve their SimBots throughout the Internal Amazon Beta Period. Sponsor may impose Availability Criteria, including but not limited to a minimum mean Interaction Rating, uptime requirements, and ability to filter offensive content, that a SimBot must meet in order to exit the Beta Program and be made available to Amazon customers through an Alexa Skill operated by Sponsor that allows users to interact with SimBots (the “Alexa Prize Skill”). Sponsor may require SimBots that have not met the Availability Criteria to remain in the Beta Program, and may set deadlines by which all SimBots must meet the Availability Criteria in order to continue participating in the Competition. Additionally, Sponsor may condition a SimBot’s inclusion in the Alexa Prize Skill on that SimBot continuing to meet the Availability Criteria.

PHASE 11 “Initial Feedback Period”: From November 30, 2022 through February 1, 2023, the SimBots that have passed certification and that have met all Availability Criteria will be made available to Alexa users through the Alexa Prize Skill, as described below. Each SimBot’s Interaction Ratings collected during the Initial Feedback Period will be provided to the Entrant Team. Entrant Teams may continue to improve their SimBots throughout the Initial Feedback Period.

The Alexa Prize Skill; Ratings. During the Internal Amazon Beta Period, Initial Feedback Period, the Quarterfinals Interaction Period, the Semifinals Interaction Period, and the Additional Feedback Period, Sponsor may allow Alexa users to invoke the Alexa Prize Skill using a particular invocation phrase. In response, Alexa may direct the user to an Entrant Team’s SimBot to interact with the user (without identifying to the user which Entrant Team’s SimBot has been selected). The SimBot will then engage the user to complete certain virtual tasks or challenges in a virtual environment. The user may converse with the SimBot until the user ends the interaction by saying “Stop” or otherwise indicating the end of the current interaction.

At the end of the interaction, users will be asked to rate the SimBot on a scale of 1 to 5 based on the user’s satisfaction with the assistance provided by the SimBot (the “Interaction Ratings”). Sponsor may determine in its sole discretion when and how frequently each SimBot is made available to users and the method by which users invoke the Alexa Prize Skill. All SimBots may not be selected to interact with users the same number of times. When calculating Interaction Ratings for any SimBot and for any purpose, Sponsor may elect to disregard Interaction Ratings from Alexa users for the purpose of selecting a finalist SimBot if (a) the number of Interaction Ratings for a SimBot from the applicable Alexa user exceeds a limit determined by Sponsor in its sole discretion (in which case Interaction Ratings will be counted up to such limit), (b) Sponsor determines in its sole discretion that any individual has submitted Interaction Ratings using more than one user account, (c) Sponsor determines in its sole discretion the Interaction Ratings may have been generated by script, macro, bot, or other automated means, or (d) Sponsor determines in its sole discretion that inclusion of an Interaction Rating would reduce the integrity of the Interaction Ratings. Any Interaction Ratings disregarded by Sponsor will be void. Entrant Teams are prohibited from rating their own or any other SimBot or from obtaining Interaction Ratings or user interactions by any fraudulent or inappropriate means, including, without limitation, paying for Interaction Ratings, agreeing to exchange Interaction Ratings with another person, offering prizes or other inducements to members of the public, using a script, macro, bot, or other automated means to generate user interactions or Interaction Ratings, or encouraging any other person or entity to engage in any of the foregoing, as determined by Sponsor in its sole discretion. Any attempt by an Entrant Team, any Entrant Team member or his or her family or friends, or any other individual to obtain Interaction Ratings or user interactions using any mechanism that violates these Official Rules, or any other kind of fraudulent or inappropriate mechanism, as determined by Sponsor in its sole discretion, may result in disqualification.

In addition to Interaction Ratings provided by users, each SimBot interaction will be evaluated on the basis of Goal Success Rate (i.e., did the SimBot correctly complete the intended task) and Cost Function Performance (i.e., how efficiently did the SimBot complete the intended task). The scoring rubric explaining how Cost Function Performance will be calculated will be shared with all Entrant Teams prior to the beginning of the Initial Feedback Phase. Entrant Teams’ SimBots will be ranked throughout the competition using an overall score comprised of a weighted average of Interaction Ratings, Goal Success Rate, and Cost Function Performance (“Overall Ranking”).

PHASE 12 “Semifinals Interaction Period”: Between February 2, 2023 and March 22, 2023, Alexa users will have the opportunity to interact with and rate eligible SimBots through the Alexa Prize Skill. Entrant Teams may continue to improve their SimBots throughout the Semifinals Interaction Period. Each SimBot’s Interaction Ratings collected during the Semifinals Interaction Period may be provided to that SimBot’s Entrant Team, as determined by Sponsor in its sole discretion. At the conclusion of the Semifinals Interaction Period, the three SimBots with the highest Overall Ranking (with ties broken by the Goal Success Rate of each SimBot) and one or more SimBots selected by Sponsor in its sole discretion based on Sponsor’s evaluation of their relevance, coherence, accuracy, interestingness, speed and scientific merit (the “Finalist SimBots”) will proceed to the Finals Interaction Period as described below.

Disclosure of Third-Party Funding. If any Entrant Team receives any third-party funding to facilitate its participation in this Competition, such funding must be disclosed to Sponsor no later than January 1, 2023, along with any requirements imposed on the Entrant Team in connection with the funding. Entrant Teams may not accept or use any third-party funding if acceptance or use of that funding, or any requirements imposed in connection with that funding, would conflict with these Official Rules.

PHASE 13 “Finals Interaction Period”: Between March 23, 2023 and May 26, 2023, Alexa users will have the opportunity to interact with and rate eligible Finalist SimBots through the Alexa Prize Skill. Entrant Teams may continue to improve their SimBots throughout the Finals Interaction Period. Each Finalist SimBot’s Interaction Ratings collected during the Finals Interaction Period may be provided to that Finalist SimBot’s Entrant Team, as determined by Sponsor in its sole discretion. The average of the Interaction Ratings collected for a SimBot during the Finals Interaction Period are the “Finals Interaction Rating Average.” At the conclusion of the Finals Interaction Period, the three Finalist SimBots with the highest Overall Ranking (with any ties broken by the Goal Success Rate of each SimBot) will be awarded prizes as described below.

Each Finalist Entrant Team agrees to maintain the SimBot so that it is available to Alexa customers for at least six months following the public announcement of the winner of the Competition. If any Finalist Entrant Team is found to be in non-compliance with these Official Rules or raises other issues of significant concern to Sponsor, the Entrant Team may be disqualified, and Sponsor shall have no further obligation to such Finalist Entrant Team. An alternate Finalist Entrant Team may be selected by Sponsor in its sole discretion, time permitting.

Technical Publication. By no later than March 22, 2023, all participating Entrant Teams must submit a technical article including (a) the technical approach for their SimBot, and (b) any comparative experiments performed by the Entrant Team and results of those experiments (a “Technical Article”). Entrant Team’s Technical Article must include sufficient detail to permit other researchers to replicate the work. However, Technical Articles may not include any Non-Public Materials, Interaction Data, or other confidential information of Sponsor or its affiliates. If a Technical Article does not provide sufficient detail to replicate the work, Sponsor may require the Entrant Team to provide Sponsor any additional information needed to replicate the work. Sponsor will publish the Technical Articles. Prior to the publication, Sponsor will not disclose Technical Articles to third parties. Entrant Teams may update their Technical Articles prior to the Finals Event.

Entrants may publish other technical articles describing their work (“Additional Articles”), but prior to submitting any such technical articles for publication, Entrant Teams must obtain Sponsor’s written approval. Additional Articles may not include any Non-Public Materials, Interaction Data, or other confidential information of Sponsor or its affiliates. Entrants must submit any Additional Articles to Sponsor for review and comment at least two weeks prior to the submission deadline and must make, prior to submission, any changes or deletions requested by Sponsor to protect confidential or other sensitive information.

Prizes:

2021 First-Place Prize (1): The winning Entrant Team will receive $500,000 U.S. dollars awarded in the form of checks divided equally among all Entrant Team members.

2021 Second-Place Prize (1): The Entrant Team that receives the second highest Composite Score will receive $100,000 U.S. dollars awarded in the form of checks divided equally among all Entrant Team members.

2021 Third-Place Prize (1): The Entrant Team that receives the third highest Composite Score will receive $50,000 U.S. dollars awarded in the form of checks divided equally among all Entrant Team members

Prize Conditions. Prizes are non-transferable except as directed by Sponsor. No prize substitutions allowed. Except where prohibited by law, all federal, state, provincial or other tax liabilities are the responsibility of the prize winners, the Sponsor will not be responsible for any tax deductions which may be necessary and Sponsor reserves the right to withhold taxes as required by law. Prize winners will be responsible for paying all costs and expenses related to the prize that are not specifically mentioned, including, but not limited to, taxes, withholdings, and any other expenses that might reasonably be incurred by the winner in receiving or using the prize. All prizes awarded will be subject to any taxes Sponsor is required by law to withhold as well as applicable sales, use, gross receipts, goods and service, or similar transaction based taxes. IF TAXES ARE APPLICABLE TO THE PRIZE(S), IT IS THE RESPONSIBILITY OF THE WINNER TO PAY TO THE APPROPRIATE AUTHORITIES. PAYMENTS TO COMPETITION WINNERS ARE SUBJECT TO THE EXPRESS REQUIREMENT THAT THE WINNER SUBMIT TO SPONSOR ALL DOCUMENTATION REQUESTED BY SPONSOR (INCLUDING FORMS W-9 OR W-8BEN AS REQUESTED BY SPONSOR) TO PERMIT COMPLIANCE WITH ALL APPLICABLE STATE, FEDERAL, LOCAL AND FOREIGN (INCLUDING PROVINCIAL) TAX REPORTING AND WITHHOLDING REQUIREMENTS. Prize winners are responsible for ensuring that the tax documentation submitted to Sponsor complies with all applicable tax laws and requirements. If a winner fails to provide the documentation or submits incomplete documentation, the prize may be forfeited and Sponsor may, in its sole discretion, select an alternate winner.

Sponsor will divide all awards that are payable to Entrant Teams evenly among all Entrant Team members and distribute accordingly. Designation as a prize winner is subject to Entrant’s proof of compliance with these Official Rules, maintaining compliance with these Official Rules and approval by the Sponsor. All details of prizes not specified herein shall be determined solely by Sponsor.

PRIVACY: Entrants and Faculty Advisors acknowledge and agree that Sponsor may collect, store, share and otherwise use personally identifiable information provided during the application process and the Competition, including, but not limited to, name, mailing address, phone number, and email address. All personally identifiable information collected is subject to and will be used in accordance with Sponsor’s Privacy Notice (www.amazon.com/privacy), Eval.ai’s Privacy Policy (https://eval.ai/privacy-policy), and YouNoodle’s Privacy Policy (www.younoodle.com/privacy), including for administering the Competition and verifying Entrants’ and Faculty Advisors’ identities, addresses and telephone numbers in the event an entry qualifies for a prize. YouNoodle’s servers are located in the United States. By participating in this Competition, Entrants and Faculty Advisors are authorizing the transfer of personal data to the United States for purposes of administering the Competition, conducting publicity about the Competition and such additional purposes consistent with Sponsor’s goals or the Competition goals. By entering the Competition, Entrants and Faculty Advisors consent to Sponsor’s and YouNoodle’s collection, and Sponsor’s use and disclosure of entrants’ personally identifiable information for these purposes. The data controller for information collected by Sponsor is Amazon.com Services LLC, 410 Terry Ave North, Seattle, Washington 98109, USA. By entering the Competition, each Entrant/Entrant Team represents and warrants that he or she has obtained all of the rights, licenses, and permissions in writing from any person who may have helped create the SimBot or that are otherwise necessary for the Entrant Team to distribute the SimBot and to grant the licenses set forth in these Official Rules and in the Developer Agreement. Entrants MAY NOT submit SimBots created by any person other than themselves or their Entrant Team members. Decisions of the Sponsor are final and binding.

INTELLECTUAL PROPERTY: As between Sponsor and Entrant Teams, all SimBots created by Entrant Teams will remain the property of the respective Entrant Teams (excluding any Program Materials incorporated in the SimBot, which will remain property of Sponsor). However, Sponsor and its designees will have rights to the SimBots as described below. By submitting a SimBot in this Competition, each Entrant represents and warrants that he or she owns, or otherwise has the right to use and license, all of the intellectual property and other rights in and to the SimBot. Each Entrant hereby grants Sponsor and its affiliates a non-exclusive, perpetual, irrevocable, worldwide, and royalty-free license to make, have made, use, sell, offer for sale, import, export, license, exploit, promote, reproduce, make available, publicly display, publicly perform, create derivative works of, and otherwise exercise all intellectual property and other rights in and to any concepts, works, inventions, information, designs, programs, or software that Entrant or his or her Entrant Team develop or submit in connection with the Competition or the creation of the SimBot, including any of the foregoing included or described in any Technical Article or other materials provided to Sponsor. In addition, upon Sponsor’s request, all Entrant Teams must provide Sponsor all source code and algorithms developed in connection with the Competition. Each Entrant agrees that he or she will do or perform, or cause to be done and performed, all such further acts and things, and shall execute and deliver all such other agreements, certificates, instruments and documents as Sponsor may reasonably request in order to carry out the intent and accomplish the purposes of the foregoing license.

Except where prohibited, each prize winner further consents (and agrees to sign any additional documents required by Sponsor to formalize, effect or perfect such consent) to Competition Entities’ use of SimBots as specified herein and the use of any Entrant names, likeness, biographical information, and voice in advertising, publicity, trade, and other marketing and promotional materials (including video, audio, and print through all means of distribution) worldwide without compensation, notice or approval, and prize winner disclaims any ownership rights to the content of such materials.

WAIVER, RELEASE, AND LIMITATION OF LIABILITY:

EACH ENTRANT ACCEPTS THE CONDITIONS STATED IN THESE OFFICIAL RULES, AGREES TO BE BOUND BY THE DECISIONS OF THE SPONSOR, WARRANTS THAT HE OR SHE IS ELIGIBLE TO PARTICIPATE IN THIS COMPETITION, AND AGREES TO RELEASE, INDEMNIFY, AND HOLD HARMLESS COMPETITION ENTITIES AND THE PERSONNEL OF EACH FROM AND AGAINST ANY AND ALL CLAIMS, LOSSES, LIABILITY, AND DAMAGES OF ANY KIND (INCLUDING REASONABLE ATTORNEYS’ FEES AND EXPENSES) ASSERTED AGAINST ANY OF THEM, INCURRED OR SUSTAINED IN CONNECTION WITH OR RISING OUT OF ENTRANT’S PARTICIPATION IN THIS COMPETITION OR ANY TRAVEL OR ACTIVITY RELATED THERETO, USE OF ANY SIMBOT OR RIGHTS THEREIN, OR BREACH OF ANY AGREEMENT OR WARRANTY ASSOCIATED WITH THE COMPETITION, INCLUDING THESE OFFICIAL RULES. ANY ATTEMPT TO DELIBERATELY DAMAGE ANY WEBSITE OR UNDERMINE THE LEGITIMATE OPERATION OF THE COMPETITION IS A VIOLATION OF CRIMINAL AND CIVIL LAWS AND, SHOULD SUCH AN ATTEMPT BE MADE, THE COMPETITION ENTITIES AND EACH OF THEIR LICENSEES RESERVE THE RIGHT TO SEEK ANY AND ALL REMEDIES AVAILABLE FROM ANY SUCH PERSON(S) RESPONSIBLE FOR ANY SUCH ATTEMPT TO THE FULLEST EXTENT PERMITTED BY LAW.

Each Entrant hereby acknowledges and agrees that the relationship between themselves and the Competition Entities is not a confidential, fiduciary, or other special relationship, and that the Entrant’s decision to provide the entry to Sponsor for purposes of the Competition does not place the Competition Entities in a position that is any different from the position held by members of the general public with regard to elements of the entry, other than as set forth in these Official Rules. Each Entrant understands and acknowledges that the Competition Entities have wide access to SimBots, technology, designs, and other materials, and that new ideas are constantly being submitted to them or being developed by their own employees. Each Entrant also acknowledges that many ideas may be competitive with, similar to, or identical to the SimBot submission in theme, idea, format, or other respects. Each Entrant acknowledges and agrees that such Entrant will not be entitled to any compensation as a result of Competition Entities' use of any such similar or identical material that has or may come to Competition Entities, or any of them, from other sources. Entrants acknowledge that other Entrants/Entrant Teams may have created ideas and concepts contained in their SimBot design that may have familiarities or similarities to their SimBot design, and that they will not be entitled to any compensation or right to negotiate with the Competition Entities because of these familiarities or similarities.

Entrants further agree that the Competition Entities are not responsible for the following: (a) electronic transmissions, SimBots or entries or notifications that are lost, late, stolen, incomplete, damaged, garbled, destroyed, misdirected or not received by Sponsor or their agents for any reason; (b) any problems or technical malfunctions, errors, omissions, interruptions, deletions, defects, delays in operation or transmission, communication failures and/or human error that may occur in the transmission, shipping errors or delays, receipt or processing of entries or related materials; or for destruction of or unauthorized access to, or alteration of, entries or related material; (c) failed or unavailable hardware, network, software or telephone transmissions, damage to Entrants’ or any person’s computer and/or its contents related to or resulting from participation in this Competition; (d) causes that jeopardize the administration, security, fairness, integrity, or proper conduct of this Competition; (e) any entries submitted in a manner that is not expressly allowed under these Official Rules (all such entries will be disqualified); or (f) any printing errors in these Official Rules or in any advertisements or correspondence in connection with this Competition or the tabulation of Interaction Ratings or scores. Sponsor reserves the right, in its sole discretion, to cancel or suspend this Competition should virus, bugs, fraud, hacking, or other causes corrupt the administration, security, or proper play of the Competition, or in the event Sponsor does not receive a minimum of two qualified entries from separate eligible Entrant Teams. Sponsor further reserves the right, in its sole discretion, to cancel or suspend this Competition or to reschedule or reformat events, including without limitation the Alexa Prize summit or Finals Event, should Sponsor be prevented, in any manner whatsoever, from holding this Competition or any event due to any present or future law (whether or not valid); any act of God, earthquake, fire, flood, epidemic (including, without limitation, any pandemic), accident, explosion or casualty; any civil disturbance or armed conflict; or any other cause of any similar nature outside of Sponsor’s control. In all such cases, notice to this effect will be posted on the Competition Site and prizes to the extent awarded will be awarded as determined by Sponsor prior to cancellation. If, in Sponsor’s opinion, there is any suspected or actual evidence of electronic or non-electronic tampering with any portion of the Competition or if technical difficulties compromise the integrity of the Competition, the Sponsor reserves the right to void suspect entries and/or terminate the Competition and award prizes in its sole discretion. Sponsor reserves the right, in its sole discretion, to disqualify any individual found tampering with the entry process or entry materials or otherwise interfering with the proper administration of the Competition or violating these Official Rules.

DISPUTES: Except where prohibited, you agree that: (1) any and all disputes, claims and causes of action arising out of or connected with this Competition or any prize awarded shall be resolved individually, without resort to any form of class action; (2) any and all claims, judgments and awards shall be limited to actual out-of-pocket costs incurred, including costs associated with entering this Competition, but in no event attorneys’ fees; and (3) under no circumstances will you be permitted to obtain awards for, and you hereby waive all rights to claim, indirect, punitive, incidental and consequential damages and any other damages (other than for actual out-of-pocket expenses), and any and all rights to have damages multiplied or otherwise increased. All issues and questions concerning the construction, validity, interpretation and enforceability of these Official Rules, or the rights and obligations of the Entrant and Sponsor in connection with the Competition, shall be governed by, and construed in accordance with, the laws of the State of Washington without giving effect to any choice of law or conflict of law rules (whether of the State of Washington or any other jurisdiction), which would cause the application of the laws of any jurisdiction other than the State of Washington.

SPONSOR: Amazon.com Services LLC, 410 Terry Ave North, Seattle, Washington 98109, USA.

Latest news

The latest updates, stories, and more about Alexa Prize.
IN, KA, Bengaluru
RBS (Retail Business Services) Tech team works towards enhancing the customer experience (CX) and their trust in product data by providing technologies to find and fix Amazon CX defects at scale. Our platforms help in improving the CX in all phases of customer journey, including selection, discoverability & fulfilment, buying experience and post-buying experience (product quality and customer returns). The team also develops GenAI platforms for automation of Amazon Stores Operations. As a Sciences team in RBS Tech, we focus on foundational ML research and develop scalable state-of-the-art ML solutions to solve the problems covering customer experience (CX) and Selling partner experience (SPX). We work to solve problems related to multi-modal understanding (text and images), task automation through multi-modal LLM Agents, supervised and unsupervised techniques, multi-task learning, multi-label classification, aspect and topic extraction for Customer Anecdote Mining, image and text similarity and retrieval using NLP and Computer Vision for product groupings and identifying duplicate listings in product search results. Key job responsibilities As an Applied Scientist, you will be responsible to design and deploy scalable GenAI, NLP and Computer Vision solutions that will impact the content visible to millions of customer and solve key customer experience issues. You will develop novel LLM, deep learning and statistical techniques for task automation, text processing, image processing, pattern recognition, and anomaly detection problems. You will define the research and experiments strategy with an iterative execution approach to develop AI/ML models and progressively improve the results over time. You will partner with business and engineering teams to identify and solve large and significantly complex problems that require scientific innovation. You will help the team leverage your expertise, by coaching and mentoring. You will contribute to the professional development of colleagues, improving their technical knowledge and the engineering practices. You will independently as well as guide team to file for patents and/or publish research work where opportunities arise. The RBS org deals with problems that are directly related to the selling partners and end customers and the ML team drives resolution to organization level problems. Therefore, the Applied Scientist role will impact the large product strategy, identifies new business opportunities and provides strategic direction which is very exciting.
US, NY, New York
We are seeking a Human-Robot Interaction (HRI) Research Scientist to develop cutting-edge interactions that make robots feel alive, personal, and fun. In this role, you will focus on verbal and non-verbal conversational systems, social dynamics, memory, and long-term relationship formation between robots, their environments, and the people they interact with. Your contributions will be essential in advancing robotics by enabling expressive, socially intelligent, and trustworthy interactions between robots and humans.
US, MA, North Reading
Amazon Robotics is transforming warehouse automation through edge AI and machine learning applied to real-world robotics challenges. We're seeking a Research Scientist to advance our mobile manipulation capabilities by developing novel learning-based approaches that enable robots to navigate and manipulate objects in dynamic fulfillment environments. This role offers the opportunity to conduct original research and translate state-of-the-art findings into production systems operating at Amazon's unprecedented scale. Key job responsibilities Research and Algorithm Development: Formulate novel research problems in robot learning and manipulation, design new model architectures, validate hypotheses through rigorous experimentation, and advance the state of the art in learning-based robotics. Data Strategy and Pipeline Design: Define data requirements for research initiatives, design scalable collection and curation strategies, establish governance and provenance standards, and build reusable pipelines ensuring data quality and reproducibility. Experimentation and Scientific Validation: Design and execute experiments in simulation and real-world embodiments, develop evaluation methodologies and benchmarks, perform ablation studies and statistical analyses, and iterate systematically to advance model performance. Prototyping and Research Infrastructure: Develop clean, well-documented research codebases, build experimentation frameworks and evaluation tooling, contribute to shared training infrastructure, and implement interfaces for broader robotics integration. Scientific Leadership and Publication: Drive an independent research agenda aligned with team objectives, publish at top-tier venues (e.g., RSS, CoRL, ICRA, NeurIPS), identify research gaps through literature reviews, and present findings via technical reports and talks. Cross-Functional Collaboration: Partner with scientists, engineers, and leaders across teams to translate research into deployable solutions, mentor junior researchers, contribute to the team's scientific culture, and support integration with robotics hardware teams. A day in the life If you are not sure that every qualification on the list above describes you exactly, we'd still love to hear from you! At Amazon, we value people with unique backgrounds, experiences, and skillsets. If you’re passionate about this role and want to make an impact on a global scale, please apply! About the team Are you inspired by invention? Is problem solving through teamwork in your DNA? Do you like the idea of seeing how your work impacts the bigger picture? Answer yes to any of these and you’ll fit right in here at Amazon Robotics. We are a smart, collaborative team of enthusiastic doers that work passionately to apply innovative advances in robotics and software to solve real-world challenges that will transform our customers’ experiences in ways we can’t even image yet. We invent new improvements every day. We are Amazon Robotics and we will give you the tools and support you need to invent with us in ways that are rewarding, fulfilling and fun!
US, NY, New York
Want to work on building a Amazon Ads billion dollar business, innovate on a new product, and have a positive impact on millions of views while working with industry-leading technologies? We're growing a team to support the Sponsored Ads business that powers the advertising experience for millions of viewers and advertisers daily. Amazon is investing heavily in building a world-class advertising business and developing a collection of self-service performance advertising products that drive discovery and sales. We deliver billions of ad impressions and millions of clicks daily and are constantly challenging ourselves to create world-class products and an unparalleled shopping experience for our hundreds of millions of customers worldwide. Key job responsibilities We are building the next-gen smart ads campaign. At its core is an Intelligence Flywheel — an architecture where every component's output is designed to train models that improve every other component. The Model Layer that is meant to power every capability currently has no dedicated science ownership. As the Senior Applied Scientist on this team, you own the science that makes the flywheel turn. You will turn static, threshold-based logic into self-improving, closed-loop intelligence, and define the decision policies that let the system act autonomously with advertiser trust. Concretely, you will: * Build predictive issue-detection models that identify under-delivery, over-delivery, and performance degradation from campaign signals before they materially impact advertisers. * Design the intervention-selection policy — which autonomous action to take — framed as a contextual bandit: choose, observe, update. Engineers implement action execution; you define the policy that selects actions. * Establish causal attribution for autonomous optimization, separating the effect of our interventions from organic performance change, so improvements can be attributed and advertiser trust in autonomy can be earned. * Integrate and adapt cross-model signals. Combine creative-quality, product-relevance, and budget signals into a unified advertiser-intelligence picture, and adapt general-purpose partner models to our product reality via fine-tuning, re-ranking, or thin adaptation layers. * Close recommendation and grading feedback loops — define reward schemas for accept/reject and performance-vs-baseline signals, correlate creative-quality scores with real campaign outcomes, and feed empirical findings back to both our product and partner science teams. You will work backwards from ambiguous business problems, set the science roadmap for the Model Layer, and partner closely with the team's software engineers — who own the services, pipelines, and execution infrastructure — so that model artifacts you produce are deployed and served in production. This is a high-leverage, high-autonomy role: your outputs are consumed by multiple engineering workstreams at once, and you set the abstractions the team builds on. About the team We are focused on goal-oriented, AI powered workflows that help advertisers achieve their marketing objectives. We collect campaign goals, surface relevant data at key decision points, and provide reporting that validates decision-making. Our product suite guides advertisers in building campaigns with optimal targeting, creative formats, inventory, and bid models that are highly likely to hit their goals — reducing the need for manual intervention.
US, WA, Seattle
At Amazon Selection and Catalog Systems (ASCS), our mission is to power the online buying experience for customers worldwide so they can find, discover, and buy any product they want. We innovate on behalf of our customers to ensure uniqueness and consistency of product identity and to infer relationships between products in Amazon Catalog to drive the selection gateway for the search and browse experiences on the website. We're solving a fundamental AI challenge: establishing product relevant information at unprecedented scale with Frontier Models and Agents. The scale is staggering: billions of products, petabytes of multimodal data, millions of sellers, dozens of languages, and infinite product diversity ranging from electronics to groceries to digital content. The research challenges are immense. GenAI and VLMs hold transformative promise for catalog understanding, but we operate where traditional methods fail: ambiguous problem spaces, incomplete and noisy data, inherent uncertainty, reasoning across both images and textual data, and explaining decisions at scale. Enriching product information requires sophisticated models that reason across text, images, and structured data, all while maintaining accuracy and trust for high-stakes business decisions affecting millions of customers daily. Amazon's Catalog System Services Science team is looking for an innovative and customer-focused applied scientist to help us make the world's best product catalog even better. In this role, you will partner with technology and business leaders to build new state-of-the-art algorithms, models, and services. You will pioneer advanced GenAI solutions that power next-generation agentic shopping experiences, working in a collaborative environment where you can experiment with massive data from the world's largest product catalog, tackle problems at the frontier of AI research, rapidly implement and deploy your algorithmic ideas at scale, across millions of customers. Key job responsibilities - Formulate novel research problems at the intersection of GenAI, multimodal learning, and large-scale information retrieval. In essence, translating ambiguous business challenges into tractable scientific frameworks - Design and implement leading models leveraging frontier models, and agentic architectures to enrich catalog information at billion-product scale - Pioneer explainable AI methodologies that balance model performance with scalability requirements for production systems impacting millions of daily customer decisions - Own end-to-end ML pipelines from research ideation to production deployment, processing petabytes of multimodal data with rigorous evaluation frameworks - Represent the team in the broader science community - publishing findings, delivering tech talks, and staying at the forefront of GenAI, VLM, and agentic system research
US, WA, Seattle
What happens when you give AI the ability to remember? Not cached responses — real structured memory that compounds over time and transfers across contexts. We're building the science behind this, and we need researchers who want to own the problem end-to-end. This is a founding role on a new team. You won't inherit models or maintain someone else's pipeline. You'll define the research direction, run experiments at scale, and ship what works directly to production. Key job responsibilities As an Applied Scientist in our team, you will be responsible for the research, design, and development of new AI technologies for knowledge acquisition and retrieval. You will adopt or invent new machine learning and analytical techniques in the realm of information retrieval, knowledge representation, and large language models. Specific responsibilities include: 1. Design and implement novel approaches to knowledge extraction from heterogeneous, unstructured data sources at organizational scale. 2. Build retrieval systems that match intent to relevant knowledge across domains — solving the "right memory at the right time" problem. 3. Own the quality of memory generation: what to capture, how to structure it, when to surface it, and when to let it decay. 4. Run large-scale experiments using Amazon's compute infrastructure and massive real-world datasets. 5. Develop evaluation frameworks for a system where "quality" means something new — right knowledge, right context, right confidence level. 6. Collaborate with engineers to move from research prototype to production system in weeks, not quarters. 7. Invent new approaches to temporal knowledge management — how memories age, conflict, and compound over time. 8. Publish and patent novel approaches to knowledge acquisition and retrieval at top-tier venues. A day in the life You will solve real-world problems by getting and analyzing large amounts of data, generate insights and opportunities, execute experiments, and develop statistical and ML models. The team is driven by business needs, which requires collaboration with other Scientists, Engineers, and Product Managers across the organization. You get to influence stakeholders with clear communication skills. You innovate on behalf of the customer and strategically build features. You will mentor junior members and help them grow. About the team We're a new team within Personalization, focused on a different kind of recommendation: not "what product should this customer see" but "what knowledge should this AI use right now." Same scale, same rigor, entirely new problem space. The science is at the intersection of information retrieval, knowledge representation, and LLM reasoning — and the right approach hasn't been established yet. The team values innovation and offers a safe place to try, fail, and learn while fostering a culture of continuous improvement. Everyone is a leader and owner for everything we do as a team. We offer creative space with an entrepreneurial work environment focusing on customer obsession.
IN, HR, Gurugram
Building large-scale forecasting and optimization systems that power Amazon’s global transportation network and directly impact customer experience and cost. Key job responsibilities 1. Guide model and system design across a range of techniques, including tree-based models, deep learning (LSTMs, transformers), LLMs, and reinforcement learning. 2. Ensure models are production-ready, scalable, and robust through close partnership with stakeholders. 3. Partner with Product, Operations, and Engineering leaders to enable proactive decision-making and corrective actions. 4 Own end-to-end business metrics, directly influencing customer experience, cost optimization, and network reliability. 5. Help contribute to the broader ML community through publications, conference submissions, and internal knowledge sharing.
US, WA, Seattle
Prime Video is a first-stop entertainment destination offering customers a vast collection of premium programming in one app available across thousands of devices. Prime members can customize their viewing experience and find their favorite movies, series, documentaries, and live sports – including Amazon MGM Studios-produced series and movies; licensed fan favorites; and programming from Prime Video subscriptions such as Apple TV+, HBO Max, Peacock, Crunchyroll and MGM+. All customers, regardless of whether they have a Prime membership or not, can rent or buy titles via the Prime Video Store, and can enjoy even more content for free with ads. Are you interested in shaping the future of entertainment? Prime Video's technology teams are creating best-in-class digital video experience. As a Prime Video team member, you’ll have end-to-end ownership of the product, user experience, design, and technology required to deliver state-of-the-art experiences for our customers. You’ll get to work on projects that are fast-paced, challenging, and varied. You’ll also be able to experiment with new possibilities, take risks, and collaborate with remarkable people. We’ll look for you to bring your diverse perspectives, ideas, and skill-sets to make Prime Video even better for our customers. With global opportunities for talented technologists, you can decide where a career Prime Video Tech takes you!
US, CA, Sunnyvale
We are looking for a Senior Applied Scientist to help drive the research and development of real-time multimodal conversational AI. You will contribute across two focus areas: advancing foundation models for speech and audio, and building the post-training systems (reward modeling, reinforcement learning) that shape natural, human-like conversational behavior. You will own a significant research area and contribute across the full model lifecycle — from pre-training and architecture design through post-training alignment and real-time deployment. You will work at the frontier of what’s possible in conversational AI, with the compute, data, and runway to pursue problems that few teams in the world have the resources to tackle. As a Senior Scientist, you will drive the technical execution of your research area, contribute to the team’s roadmap, and work closely with inference engineers to ensure your models are designed for real-time production deployment. Key job responsibilities What You’ll Do Foundation Model Scaling - Help build and train large-scale multimodal foundation models for real-time speech and audio generation, from architecture design through production-scale training - Advance the scaling and efficiency of conversational models, including the relationship between data, model size, and real-time performance - Design model architectures informed by hardware constraints and inference requirements, working with inference engineers to ensure models are servable from inception - Develop training methodologies for multimodal models that jointly process and generate speech, language, and audio in real-time streaming contexts - Contribute to the state of the art on efficient architectures and training methods for conversational AI at scale Post-Training & Reinforcement Learning - Design and build reward models and reward functions for speech systems — capturing naturalness, fluency, conversational quality, and real-time responsiveness - Develop and apply reinforcement learning methods to shape conversational behavior — teaching models natural timing, responsiveness, and fluid interaction - Build parts of the post-training pipeline from SFT through RL alignment, optimized for real-time multimodal outputs rather than text-only generation - Design evaluation frameworks that capture the quality dimensions unique to real-time conversation (latency sensitivity, audio quality, prosody, interaction naturalness) Real-Time Perception & Generation - Advance the team’s capabilities in real-time perception — the ability of the model to process incoming audio/speech while simultaneously generating responses - Develop techniques for natural interactive systems where the model handles concurrent input and output with human-like timing - Work at the intersection of model architecture and production constraints to ensure multimodal capabilities function within hard real-time latency budgets
ES, B, Barcelona
How does Amazon decide which fulfillment center ships your order, which truck carries it, and how to keep promises across hundreds of millions of packages daily? How does it decide how many trucks and how much labor are required to ship orders across the network? SCOT Fulfillment Optimization (FO) owns the optimization and forecasting science behind these decisions. We are seeking Applied Scientists to join the FO Science & Tech team in Barcelona (alternatively: Luxembourg or London) with a strong academic background in optimization, machine learning, and/or time-series forecasting. • You will design and build state-of-the-art machine learning and optimization models that power Amazon's fulfillment decisions at an unprecedented scale across two core scientific pillars: • Large-Scale Optimization and Planning: Designing planning systems for order assignment and resource utilization, while balancing multi-objective cost-speed tradeoffs to enable controllers to steer millions of shipments per hour optimally. • Demand Forecasting & Predictive ML: Developing time-series forecasts for customer demand, incorporating contextual information (weather, sales, order properties), and modeling uncertainty for core planning systems. Basic qualifications • PhD in Operations Research, Applied Mathematics, Computer Science, or related field (or equivalent experience) • Strong programming skills (Python preferred; experience with optimization solvers a plus) • Research experience in one or more: • Large-scale mathematical programming (LP, MIP, decomposition methods) • Combinatorial optimization (assignment, scheduling, network flows) • Multi-objective optimization and control • Large-scale time-series forecasting (GenAI models, probabilistic forecasting, uncertainty quantification) • Causal inference (spatiotemporal causal modeling, offline policy evaluation) Preferred qualifications • Experience building optimization systems that run in production at scale • Being comfortable with ambiguity and fast iteration cycles • Publications in relevant venues Key job responsibilities Design and implement optimization and forecasting models for large-scale fulfillment problems, from order assignment to network flow control. Build research prototypes end-to-end: from problem formulation through scalable implementation to production validation. Analyse complex tradeoffs (cost, speed, capacity, accuracy) and translate findings into actionable recommendations for leadership and operations teams. Collaborate with engineers to bring science solutions into production systems serving millions of customer orders daily. A day in the life You formulate an optimization or forecasting problem on a whiteboard with teammates, then prototype it in Python with real data by the afternoon. You run experiments against production-scale datasets, iterate on the model, and present results to stakeholders who will use them to make network decisions next week. Some days you dive deep into solver performance; other days you're explaining a Pareto frontier to an operations leader. You collaborate with large engineering and product teams to bring your solutions into systems serving millions of customers. Alongside fast-turnaround prototypes, you own long-term research bets, the kind that reshape how Amazon's fulfillment network operates at scale. Your work goes live. About the team SCOT Fulfillment Optimization Science & Tech (FO SnT) is the applied research team behind Amazon's fulfillment decision-making systems. We decide how orders get assigned to warehouses, how capacity is allocated across the network, and how cost and speed tradeoffs are managed in real time, at global scale. Our models influence billions of euros in annual operational spend. They protect sites from overload during peak, reduce transportation costs and CO2 emissions, and ensure customers receive their packages when promised. Leadership relies on our science to make investment decisions worth hundreds of millions. We are practitioners of large-scale optimization: MIP formulations, decomposition methods, approximation algorithms, and parallelisation. We use machine learning where it sharpens our decisions, including forecasting, learned heuristics, and multi-armed bandits. We pick the right tool for the problem, not the fashionable one. You will work alongside Senior and Principal scientists, and collaborate with Amazon Scholars and academic partners who bring frontier research into our applied problems. We code our prototypes to be production-ready and collaborate with large engineering teams to ship systems, not papers. Above all, we have fun solving hard real-world problems at real-world speed, failing, learning, and shipping along the way.