Machine-labeled data + artificial noise = better speech recognition

Although deep neural networks have enabled accurate large-vocabulary speech recognition, training them requires thousands of hours of transcribed data, which is time-consuming and expensive to collect. So Amazon scientists have been investigating techniques that will let Alexa learn with minimal human involvement, techniques that fall in the categories of unsupervised and semi-supervised learning.

At this year’s International Conference on Acoustics, Speech, and Signal Processing, my colleagues and I are presenting a semi-supervised-learning approach to improving speech recognition performance — especially in noisy environments, where existing systems can still struggle.

We first train a speech recognizer — the “teacher” model — on 800 hours of annotated data and use it to “softly” label another 7,200 hours of unannotated data. Then we artificially add noise to the same dataset and use that, together with the labels generated by the teacher model, to train a second speech recognizer — the “student” model. We hope to make the behavior of the student model in the noisy domain approach that of the teacher model in the clean domain, and thus improve the noise robustness of the speech recognition system.

The architecture of our teacher-student model. "Logits selection" refers to the selection of high-confidence senones.

On test data that we produced by simultaneously playing recorded speech and media sounds through loudspeakers and re-recording the combined acoustic signal, our system shows a 20% relative reduction in terms of word error rate versus a system trained only on the clean, annotated data.

An automatic speech recognition system has three main components: an acoustic model, a pronunciation model, and a language model. The inputs to the acoustic model are short snippets of audio called frames. For every input frame, the output is thousands of probabilities. Each probability indicates the likelihood that the frame belongs to a low-level phonetic representation called a senone.

In training the student model, we keep only the highest-confidence senones from the teacher, which turns out to be a quite effective approach.

The outputs of the acoustic model pass to the pronunciation model, which converts senone sequences into possible words, and those pass to the language model, which encodes the probabilities of word sequences. All three components of the system work together to find the most likely word sequence given the audio input.

Both our teacher and student models are acoustic models, and we experiment with two criteria for optimizing them. With the first, the models are optimized to maximize accuracy on a frame-by-frame basis, at the level of the acoustic model. The other training criterion is sequence-discriminative: both the teacher and student models are further optimized to minimize error across sequences of outputs, at the levels of not only the acoustic model but the pronunciation model and language model as well.

We find that sequence training makes the teacher models more accurate, apart from the performance of the student models. It also slightly increases the relative improvement offered by the student models.

To add noise to the training data, we used a collection of noise samples, most of which involved media playback — such as music or television audio — in the background. For each speech example in the training set, we randomly selected one to three noise samples to add to it. Those samples were processed to simulate closed-room acoustics, with the properties of the simulated room varying randomly from one training example to the next.

For every frame of audio data that passes to an acoustic model, most of the output probabilities are extremely low. So when we use the teacher’s output to train the student, we keep only the highest probabilities. We experimented with different numbers of target probabilities, from five to 40.

Intriguingly, this modification by itself improved the performance of the student model relative to the teacher, even on clean test data. Training the student to ignore improbable hypotheses enabled it to devote more resources to distinguishing among probable ones.

In addition to limiting the number of target probabilities, we also applied a smoothing function to them, which evened them out somewhat, boosting the lows and trimming the highs. The degree of smoothing is defined by a quantity called temperature. We found that a temperature of 2, together with keeping the 20 top probabilities, yielded the best results.

Apart from the data set produced by re-recording overlapping audio, we used two other data sets to test our system. One was a set of clean audio samples, and the other was a set of samples to which we’d added noise through the same procedure we used to create the training data.

Our best-performing student model was first optimized according to the per-frame output from the teacher model, using the entire 8,000 hours of data with noise added, then sequence-trained on the 800 hours of annotated data. Relative to a teacher model sequence-trained on 800 hours of hand-labeled clean data, it yielded a 10% decrease in error rate on the clean test data, a 29% decrease on the noisy test data, and a 20% decrease on the re-recorded noisy data.

Acknowledgments: Ladislav Mosner, Anirudh Raju, Sree Hari Krishnan Parthasarathi, Kenichi Kumatani, Shiva Sundaram, Roland Maas, Björn Hoffmeister

About the Author
Minhua Wu is an applied scientist in the Alexa Speech group.

Related content

Work with us

See more jobs
US, NY, New York City
Job summaryAmazon is the 4th most popular site in the US ( Our product search engine is one of the most heavily used services in the world, indexes billions of products, and serves hundreds of millions of customers world-wide. We are working on a new AI-first initiative to re-architect and reinvent the way we do search through the use of extremely large scale next-generation deep learning techniques. Our goal is to make step function improvements in the use of advanced Machine Learning (ML) on very large scale datasets, specifically through the use of aggressive systems engineering and hardware accelerators. This is a rare opportunity to develop cutting edge ML solutions and apply them to a problem of this magnitude. Some exciting questions that we expect to answer over the next few years include:· Can a focus on compilers and custom hardware help us accelerate model training and reduce hardware costs?· · Can combining supervised multi-task training with unsupervised training help us to improve model accuracy?· · Can we transfer our knowledge of the customer to every language and every locale ?This is a unique opportunity to get in on the ground floor, shape, and build the next-generation of Amazon Search. We are looking for exceptional scientists and ML engineers who are passionate about innovation and impact, and want to work in a team with a startup culture within a larger organization.Please visit for more information
US, WA, Seattle
Job summaryMultiple positions availableCompany: Services LLCLocation: Seattle, WA1. Participate in the design, development, evaluation, deployment and updating of data-driven models and analytical solutions for machine learning and/or natural language applications.2. Develop and/or apply statistical modeling techniques, optimization methods, and other ML techniques to different applications in business and engineering.3. Routinely build and deploy ML models on available data.4. Research and implement novel ML and statistical approaches to add value to the Business.5. Collaborate with other teams to find technical solutions to complex problems.6. Mentor junior engineers and scientist.A day in the life
US, WA, Seattle
Job summaryAre you interested in big data, machine learning, and product recommendations? If so, the Product Semantics team in Amazon Product Graph might be the right place for you. We are a team in a fast-paced organization with a huge impact on hundreds of millions of customers. We innovate at the intersection of customer experience, deep learning, and high-scale machine-learning systems.As the world’s most customer-centric company, Amazon heavily invests in inventing and applying state-of-art technologies to build world-class product recommendation systems to improve shopper experience. We break fresh ground to create world-class customer-facing features to help customers discover high quality products that meet their needs, and provide most relevant product information to help customers make confident shopping decisions. We are a highly motivated, collaborative, and fun-loving team with a strong entrepreneurial spirit and bias for action. With a broad mandate to experiment and innovate, we have a very wide range of new opportunities to explore.The Product Semantics team in Amazon Personalization, based in Seattle and New York City, is looking for scientists who love big data, are passionate about understanding products and product relationships from product profiles, reviews, and search log, and who are capable of inventing and applying Machine Learning, NLP, and Computer Vision techniques that will leave no valuable data behind. Our applied scientists work closely with software engineers to put algorithms into practice. They also work in partnership with teams across Amazon to create enormous benefits for our customers.If you have an entrepreneurial spirit, know how to deliver, love to work with data, are deeply technical, highly innovative and long for the opportunity to build solutions to challenging problems that directly impact the company's bottom-line, we want to talk to you.Key job responsibilities· Use machine learning and analytical techniques to create scalable solutions for business problems· Analyze and extract relevant information from large amounts of Amazon's historical business data to help automate and optimize key processes· Design, development and evaluation of highly innovative models for predictive learning· Work closely with software engineering teams to drive model implementations and new feature creations· Establish scalable, efficient, automated processes for large scale data analyses, model development, model validation and model implementation· Research and implement novel machine learning and statistical approachesAbout the teamOur mission is to delight every Amazon customer with a personalized shopping experience. We achieve our mission through investments in UX, Science, and Systems with the purpose of delivering the future of shopping on Amazon. We are seeking an Applied Scientist to work on step function science improvements across the recommendations space.
US, VA, Arlington
Job summaryMultiple positions availableCompany: Services LLCLocation: Arlington, VA1. Participate in the design, development, evaluation, deployment and updating of data-driven models and analytical solutions for machine learning and/or natural language applications.2. Develop and/or apply statistical modeling techniques, optimization methods, and other ML techniques to different applications in business and engineering.3. Routinely build and deploy ML models on available data.4. Research and implement novel ML and statistical approaches to add value to the Business.5. Collaborate with other teams to find technical solutions to complex problems.6. Mentor junior engineers and scientist.A day in the life
LU, Luxembourg
Job summaryAre you a talented and inventive scientist with strong passion about using theoretical data science model in an applied environment? Would you like to play a key role within EU RME Predictive Analytics team? Our mission is to provide EU RME with the technical expertise to support World Class Maintenance and Spare Parts Programs. As Data Scientists you will be working with large distributed systems of data and providing Machine Learning (ML) and Predictive Modeling expertise for over 2000 maintenance engineers, managers and administrators by supporting the entire network managed by EU RME, which may include non-EU locations (such as Singapore, Australia and Japan). You will connect with world leaders in your field and you will be tackling ML challenges by carrying out a systematic review of existing solutions. The appropriate choice of the methods and their deployment into effective tools will be the key for the success in this role.The successful candidate will be a self-starter comfortable with ambiguity, with strong attention to detail and outstanding ability in balancing technical leadership with strong business judgment to make the right decisions about model and method choices.Job responsibilities· Provide technical expertise to support team strategies that will take EU RME towards World Class predictive maintenance practices and processes, driving better equipment up-time and lower repair costs with optimized spare parts inventory and placement;· Provide technical expertise to support the development of long-term spares management strategies that will ensure spares availability at an optimal level for local sites and reduce the cost of spares;· Provide technical expertise to support the development of predictive and optimization models used to reduce energy consumption and promote sustainability;· Collaborate with EU RME internal and external stakeholders and have a cross-team impact;· Create and share with audiences of varying levels technical papers and presentation.
US, WA, Seattle
Job summaryThe Amazonian Experience and Technology (AET) is in charge of providing the best experience possible to Amazon employees when interacting HR. To do so, we develop tools, technologies, and processes to ensure that Amazon can fulfill its goal of becoming Earth's Best Employer. As an Applied Scientist in this role, you will be in charge of developing state of the art Machine Learning models and developing research to enhance our ability to serve our customers.We are looking for passionate Applied Scientist with an interest in Natural Language Processing and data analytics. In this role, you will work with service owners across the globe to help automate or improve manual processes to help us solve employee problems. You will come across challenging problems, using data to solve issues such as document and image classification, automatic translation, intelligent routing, and text autocompletion that will be used by employees and AET agents to improve the experience of Amazonians reaching out to HR.About the teamThe Data, Analytics and Science, Research and Applied Science team provides rigorous scientific support to business partners across the globe within the Amazonian Experience and Technology organization. Our goal is to make operations as efficient as possible by leveraging science to make sure the experience of Amazonians when contacting HR is as frustration free as possible. In this team, you will work with Economists, Applied Scientists, Data Scientists and Software Development Engineers to help solve complicated problems for Amazon's ever-growing employee population.
US, WA, Seattle
Job summaryDo you want to join an innovative team of scientists who invent and apply the most advanced machine learning, NLP and machine translation techniques to create the best customer engagement experience on the earth? Do you want to revolutionize the way how customers solve their issues and got their questions answered? Do you want to help enabling any Amazon associates to aid any Amazon customers no matter what language they speak? At Customer Engagement Technology, we develop peculiar products that help customers solve problems. Our team leads the technical innovations in these spaces and set the bar for every other company that exists. We love data, and we have more than anyone else in the industry. We innovate on behalf of customers, developing Bot, self-service, and associate-facing products that delight customers and support our world class customer service workforce. We leverage big data, NLP, ML, and a focus on continuous innovation to create an amazing experience for customers as we scale to meet business growth each year.If you like to own solving end-to-end business problems with machine learning which would have a direct impact on the bottom line of Amazon’s business while improving customer experience, if you see how big data and cutting-edge technology can be used to improve customer experience, if you love to innovate, to discover knowledge from big structured and unstructured data and if you deliver results, then we want you to be in our team.Major responsibilities· Use statistics, NLP and machine learning techniques to create scalable solutions for business problems· Analyze and extract relevant information from large amounts of both structured and unstructured data to help automate and optimize key processes· Design, experiment and evaluate highly innovative models for predictive learning· Work closely with software engineering teams to drive real-time model experiments, implementations and new feature creations· Work closely with business staff to optimize various business operations· Establish scalable, efficient, automated processes for large scale data analyses, model development, model validation and model implementation· Track general business activity and provide clear, compelling management reporting on a regular basis· Research and implement novel machine learning and statistical approaches
US, WA, Virtual Location - Washington
Job summaryThe Worldwide Workplace Health and Safety (WHS) team is hiring a Data Scientist to help us analyze, process, and model data to create actionable plans to measure and improve Amazon’s organizational culture as it relates to health and safety of our employees. The role will analyze the impact of behavioral and organizational changes, support data-driven decision making by business leaders, and facilitate the development of innovative products that improve the safety outcomes and overall employee experience across our global operations.The role will collaborate with internal technical, business WHS and operations teams, to support the development, implementation and ongoing measurement of safety programs and initiatives to meet our vision of being Earth's safest place to work.Key job responsibilities* Support the development of start-to-finish data product solutions from requirements gathering and ideation, through interface design and implementation.* Contribute to the design and implementation of data infrastructure and pipelines for machine learning and analytics products.* Obtain, merge, analyze, and report data using SQL, statistics software, and data visualization tools.* Apply various statistical and machine learning techniques to analyze large and complex data sets related to safety engagement, safety leadership and other behavioral factors.* Communicate applied machine learning and statistic concepts to project sponsors, business leaders, and development teams across Amazon.* Understand business customer needs, iterate on feedback, and drive adoption.
US, CA, Santa Clara
Job summaryAmazon aims to exceed the expectations of our customers by ensuring that their orders, no matter how large or small, are delivered as quickly, accurately, and cost effectively as possible. To meet this goal, Amazon has invested in Amazon Logistics, a world class last mile operation.We are looking for a dynamic, resourceful, and organized Senior Data Scientist within Amazon Logistics’ Amazon Flex organization to develop new, data-driven solutions to support the most critical components of this rapidly scaling operation. The Data Scientist will work closely with other research scientists, machine learning experts, and economists to design and run experiments, research new algorithms, and find new ways to improve customer experience. The Scientist will partner with technology and product leaders to solve business and technology problems using scientific approaches to build new services that surprise and delight our customers. Science at Amazon is a highly experimental activity, although theoretical analysis and innovation are also welcome. Our scientists work closely with software engineers to put algorithms into practice. They also work on cross-disciplinary efforts with other scientists within Amazon.
LU, Luxembourg
Job summaryHave you ever ordered a product on Amazon and when that box with the smile arrived you wondered how it got to you so fast? Have you wondered where it came from and how much it cost Amazon to deliver it to you?If so, the EU Supply Chain Science team is for you. The team, based in our EU HQ in Luxembourg, is responsible for optimizing our Supply Chain, tackling problems such as inventory management and placement, warehouse flows optimization or demand forecasting.We are looking for a Senior Operations Research Scientist who will be responsible to optimize the planning of our European Fulfillment Network flows. The position will leverage ML, Optimization and Statistics methodologies to develop new mathematical models to optimize our flows across our Pan-European Network, maximizing our Customer Experience and minimizing our cost and carbon footprint. The Senior OR Scientist will also develop models to improve our risk management in Supply Chain planning and execution.You will own the models from conception to production, and will partner with technical, business or operational stakeholders to design scalable solutions to be used in production.You will also collaborate with other scientists (internal and external to the team), to share best practice or stay up-to-date on state of the art techniques.You will work in a fast-paced and ambiguous environment that requires you to be detail-oriented, fast-learning and comfortable in working with multiple operations, business and technical teams.Key job responsibilitiesYour responsibilities will include but not limited to:· Design and develop mathematical models to optimize inventory placement and product flows.· Design and develop statistical and optimization models for planning Supply Chain under uncertainty.· Manage several, high impact projects simultaneously.· Consult and collaborate with business and technical stakeholders across multiple teams to define new opportunities to optimize our Supply Chain.· Communicate data-driven insights and recommendations to diverse senior stakeholders through technical and/or business papers.· Mentor and provide feedback with and insightful reviews to peer scientists.
US, WA, Seattle
Job summaryAre you experienced at applying machine learning to big-data tasks? Are you excited by analyzing and modeling terabytes of text, images, and other types of data to solve real-world problems? We love data and we have lots of it. Join a high impact innovative team of scientists, economists and engineers who use Machine Learning, Statistics and Econometrics to develop highly innovative and impactful products that influence the company’s bottom line.The Core AI team supports the executives to drive the global optimization across almost all key business units of Amazon, including market design, pricing, forecasting, online advertising, search, supply chain network planning, and other areas. Our team is building a number of highly innovative and impactful products, such as Amazon Economics Intelligence Service, Amazon Search/Discovery optimization, and Amazon Corporate Bridging Service. Join us to build revolutionary products that has significant impact of the company’s footprint. We are looking for versatile and passionate scientists who want to develop industry leading technologies and set the bar for every other company.As an applied machine learning scientist, you work with a team of scientists to drive technology and research directions. You will work on projects of large opportunities. You will collaborate with economists, engineers and product teams. Your work will have a direct impact on the bottom line of our business while improving customer experience. If big data, cutting edge technologies and building intelligent systems excite you, if you love to innovate and deliver results, then we want you to be on our team.Key job responsibilities· Use machine learning and statistics techniques to create scalable solutions for business problems· Analyze and extract relevant information from large amounts of both structured and unstructured data including texts and images to help automate and optimize key processes· Design, experiment and evaluate highly innovative models for pricing, forecasting, natural language understanding and causal inference· Work closely with software engineering teams to drive real-time model experiments, implementations and new feature creations· Work closely with business staff to optimize various business operations· Establish scalable, efficient, automated processes for large scale data analyses, model development, model validation and model implementation· Track general business activity and provide clear, compelling management reporting on a regular basis· Research and implement novel machine learning, statistical and econometrics approaches
IL, Tel Aviv
Job summaryYou: Alexa, I am looking for a new career opportunity, where I could conduct applied research, impact millions of customers, and publish about it in top conferences. What do you suggest?Alexa: The Alexa Shopping team is looking for brilliant applied researchers to help me become the best personal shopping assistant. Do you want to hear more?You: Yes, please!Alexa: As an applied researcher in the Alexa Shopping Research team, you will be responsible for research, design, and implementation of new AI technologies for voice assistants. You will collaborate with scientists, engineers, and product partners locally and abroad. Your work will inventing, experimenting with, and launching new features, products and systems. Ideally you have a expertise in at least one of the following fields: Web search & data mining, Machine Learning, Natural Language Processing, Computer Vision, Speech Processing or Artificial Intelligence, with both hands-on experience and publications at top relevant academic venues.
ES, Madrid
Job summaryAre you excited to help customers discover the hottest and best reviewed products?Through the enablement of intelligent campaigns that leverage machine-learning models, you will help to deliver the best possible shopping experience for Amazon’s customers all over the globe.We are looking for experienced scientist who will work with business leaders, scientists, and engineers to translate business and functional requirements into concrete deliverables. Your domain spans the design, development, testing, and deployment of data driven and highly scalable solutions using data processing and machine learning in product recommendation. You will partner with scientists, product managers, and engineers to help invent and implement scalable Data processing and ML models while inventing tools on our customers behalf.A day in the lifeThis is a unique, high visibility opportunity for someone who wants to have business impact, dive deep into large-scale problems, and work closely with scientists and engineers. We are particularly interested in candidates with experience building large scale machine learning solutions and working with distributed systems to 1) help us build robust ensemble of ML systems that can drive classification and recommendation of products with a high precision and recall utilizing various signals and scale to new marketplaces and languages and 2) design optimal or near optimal supervised and unsupervised machine learning models and solutions for moderately complex projects in business, science, or engineering.About the hiring groupThe Discovery Tech team helps customers discover and engage with new, popular and relevant products across Amazon worldwide. We do this by combining technology, science, and innovation to build new customer-facing features and experiences alongside cutting edge tools for marketers. You will be responsible for creating and building critical services that automatically generate, target, and optimize Amazon’s cross-category marketing and merchandising.Job responsibilitiesAs a Senior Applied Scientist, you bring business and industry context to science and technology decisions. You set the standard for scientific excellence and make decisions that affect the way we build and integrate algorithms. Your solutions are exemplary in terms of algorithm design, clarity, model structure, efficiency, and extensibility. You tackle intrinsically hard problems, acquiring expertise as needed. You decompose complex problems into straightforward solutions.
US, WA, Seattle
Job summaryDo you want to help guide the future of Amazon’s next generation of fulfilment services to create the best customer experience by pushing the boundaries of economic modeling and experiments? The Multi Channel Fulfillment team is looking for an Economist to play a key role in guiding the next fulfillment delivery platforms to serve the fast growing needs of our merchants and Customers. Economists at Amazon are expected to develop new techniques to process large data sets, address challenging quantitative problems, guide business decisions, and contribute to the design of automated systems around the company. Upon joining the team, you will guide our businesses by building econometric models, using our world class data systems and applying economic theory to solve problems in a fast moving environment.We are looking for an Economist who will lead the development of econometric models, experiments, and algorithms that will enhance our understanding of merchant and Customer behavior, and the impact of fee and policy changes globally. The ideal candidate will have experience with large scale marketplace behavior modeling, forecasting, outstanding technical abilities, and strong business acumen. They will be comfortable with ambiguity and will enjoy working in a fast-paced dynamic environment. The position will work closely with engineers, scientists, financial analysts, and product managers.Responsibilities:· Design and develop complex and innovative models to predict the impact of fees on third party seller’s behavior, growth, and commitment· Design and conduct experiments to validate theories, and improve understanding of Amazon’s third party ecosystem and the impact changes have on the customer experience· Refine fee structures to improve the benefits for customers and sellers· Collaborate with product managers, data scientists and software developers to incorporate models into production processes and influence senior leaders· Explain complex technical approaches to a variety of stakeholders, including senior leadership, software development teams, business/product teams, and cross-functional experts· Serve as subject matter expert for economics related issues within the Multi Channel Fulfillment team
US, WA, Seattle
Job summaryThe principal scientist will initiate and work on key initiatives for developing and advancing world-leading automatic speech recognition (ASR) technology for any voice-driven Alexa end-point. The goal is to achieve unmatched speech recognition accuracy for any device, in any acoustic environment, for any speaker, and for any domain and application running on Alexa. You will analyze system short-comings, for leading the development of data-driven and algorithmic improvements, for defining the path to production, and for influencing design and architecture of goal-relevant software. You will work in a hybrid, fast-paced organization where scientists and engineers work jointly together and drive improvements directly to production.The principal scientist will either go deep on a specific area like single ASR model recognizing multiple languages supporting in-utterance code switching, or models learning without human transcription and act as a technical lead, or will work across teams and areas influencing data, algorithm, and design decisions. Areas of interest cover the whole ASR spectrum, including general purpose ASR, multi-channel raw audio input acoustic modeling, noise robust acoustic modeling, device and speaker independent acoustic modeling, acoustic model adaptation, advanced deep learning for acoustic and language modeling, active learning and semi-/unsupervised learning techniques for acoustic and language modeling, learning from heterogeneous and mismatched audio and text data including data selection and data simulation, large-scale open-domain language modeling, language model adaptation, contextual and personalized language modeling, multi-lingual automatic pronunciation generation, text verbalization and (inverse) text normalization, etc.The principal scientist will help drive scalable, robust, and automated solutions, making new algorithms and processes scalable to work on production-scale data sizes and achieving automated adaptation of processes and algorithms to new environments and to other locales. You will also help integrate new algorithms and processes into existing modeling stacks, simplify and streamline the existing modeling stacks, and develop testing and evaluation strategies. You will influence design and architecture of software stacks used offline and at runtime for building and deploying ASR model artifacts, achieving flexible yet efficient solutions suitable for R&D work and for running in production.
US, WA, Seattle
Are you seeking an environment where you can drive innovation? Do you want to apply learning techniques and advanced mathematical modeling to solve real world problems? Do you want to play a key role in the future of Amazon's Retail business? This job for you! The Customer Behavior Analytics (CBA) team, in partnership with the F3(Fresh/Food/Fast) team, is looking for motivated individuals with strong analytical skills to help drive the expansion of F3 business globally. The CBA team is responsible for the architecture, design, implementation of tools used to understand customer behavior and value generation for F3 investments. Come and join us!Amazon’s CBA team is looking for Applied Scientists, who can work at the intersection of machine learning, statistics and economics; and leverage the power of big data to solve complex problems like long-term causal effect estimation.As an applied scientist, you will bring statistical modeling and machine learning advancements to analyze data and develop customer-facing solutions in complex industrial settings. You will be working in a fast-paced, cross-disciplinary team of researchers who are leaders in the field. You will take on challenging problems, distill real requirements, and then deliver solutions that either leverage existing academic and industrial research, or utilize your own out-of-the-box pragmatic thinking. This role requires a pragmatic technical leader comfortable with ambiguity, capable of summarizing complex data and models through clear visual and written explanations. The ideal candidate will have experience with machine learning models and causal inference. Additionally, we are seeking candidates with strong rigor in applied sciences and engineering, creativity, curiosity, and great judgment.Your responsibilities include:· Understand and mine the large amount of data, prototype and implement new learning algorithms and prediction techniques to improve long-term causal estimation approaches.· Collaborate with product managers and engineering teams to design and implement solutions for Amazon problems· Design, build, and deploy effective and innovative ML solutions to improve various components of our ML and causal inference pipelines· Publish and present your work at internal and external scientific venues in the fields of ML and causal inference.Your benefits include:· Working on a high-impact, high-visibility product, with your work improving the experience of millions of customers.· The opportunity to use (and innovate) state-of-the-art ML methods to solve real-world problems.· Excellent opportunities, and ample support, for career growth, development, and mentorship.
GB, London
Job summaryAmazon is seeking an outstanding Data Scientist to uncover key insights on how customers engage with live sports events on Prime Video globally. With prestigious sporting events on Prime Video including the English Premiere League (UK), UEFA Champions League (Italy, Germany), Ligue 1 (France), US Open Tennis (UK), Roland Garros (France), Autumn Nations Cup Rugby (UK) and more, live sports are an integral and growing component of Prime Video. As our selection of events expands, the Prime Video Content Analytics team are looking to enable agile decision making on live sports by developing key insights into customer engagement with live sport and translating these insights into large scale predictive modelling and analytics solutions.You will have the following responsibilities within the scope of our global Prime Video business:· Drive analytics in an uncharted field that is developing at a fast pace but also becoming increasingly important to Prime Video· Support the analytical needs of stakeholders in the sports, advertising, finance and live events teams, inclusive of statistical inferences, demand modelling, feature engineering· Build profitability models for new sports rights and partner with finance on business cases· Think outside the box to use novel data and methodological approaches· Create new metrics that effectively guide the business and deploy dashboards to surface them to senior leadership· Ensure that the quality and timeliness of analytic deliverables meet business expectations
US, WA, Seattle
Job summaryAre you excited at the prospect of conducting research to improve the employee and manager experience at Amazon? Would you like to see your findings have a real impact on key HR programs and processes? Join the Talent Management team that supports the Advertising, IMDb, Grand Challenge (AIGC) organization (one of the fastest growing businesses at Amazon).We are seeking a Senior Research Scientist with expertise in mixed-methods research, preferably in social science and behavioral research, but are also open to experience in public health, economics, or similar fields. In this role, you will lead and support research efforts within the Recruiting, Talent Management, and Leadership & Development space.You will help set the direction for science and research in the organization. You will be a thought leader on the team, partnering with a diverse set of stakeholders to identify and develop impactful areas for novel research about talent and recruiting outcomes, mechanisms, and programs. You will mentor and provide scientific expertise/peer review to other scientists and analysts on the team.The ideal candidate should have strong problem solving skills, excellent business acumen, as well as an expertise in both qualitative and quantitative methods. This role will need to navigate complex and ambiguous business challenges by asking the right questions, understanding what methodologies to employ, and communicating results to multiple audiences (e.g., technical peers, functional teams, business leaders). Translating business and stakeholder needs into realistic and actionable scientific research will be a regular challenge in this role.Responsibilities include:· Leading scientific direction, providing consult, mentorship, and peer review.· Partnering closely and driving effective collaborations across multi-disciplinary science, analytics, HR, and business teams. Reviewing and scoping research requests, and recommending appropriate scientific methodologies.· Designing, developing, and executing qualitative and quantitative data collection, research, and experimentation. Strong experience with statistical analysis required. Experience with survey development and experimental design (including quasi-experimental) preferred.· Communicating findings and business impact effectively (written and verbally) with both technical and non-technical stakeholders.
US, CA, San Francisco
Job summaryMultiple positions availableCompany: Services LLCLocation: San Francisco, CA1. Participate in the design, development, evaluation, deployment and updating of data-driven models and analytical solutions for machine learning and/or natural language applications.2. Develop and/or apply statistical modeling techniques, optimization methods, and other ML techniques to different applications in business and engineering.3. Routinely build and deploy ML models on available data.4. Research and implement novel ML and statistical approaches to add value to the Business.5. Collaborate with other teams to find technical solutions to complex problems.6. Mentor junior engineers and scientist.A day in the life
US, WA, Seattle
Job summaryDuring the stay-at-home order, how has your grocery shopping behavior changed? Have you tried various food delivery services to feed yourself and your loved ones? The pandemic has accelerated customers' adoption of online grocery shopping services like Amazon Fresh.My organization is looking to re-imagine the world of Grocery Supply Chain here at Amazon Fresh. Our vision is to enable any grocery operation in the world to fulfill omni-channel orders by providing them with an automated sales and operations planning (S&OP) solutions for their business. We are building automated software solutions that leverage machine learning to scale S&OP from 50 to 400+ omni-channel stores and warehouses globally. We are data-driven, build scalable and automated data pipelines which yield in accurate forecasting models to support the grocery business at Amazon.To succeed, we want economists like yourself to forge a path into the future by building models to help forecast demand (being in pandemic complicates things a little bit). Economists at Amazon will be expected to develop new techniques to process large data sets, address quantitative problems, and contribute to design of automated systems around the company. You have strong leadership qualities, great judgment, clear communication skills, and a track record of shipping great products.If this sounds intriguing, then we’d like to talk to you about a role on a new Amazon team that's tackling a set of problems requiring significant innovation and scaling.Amazon is committed to a diverse and inclusive workplace. Amazon is an equal opportunity employer and does not discriminate on the basis of race, national origin, gender, gender identity, sexual orientation, protected veteran status, disability, age, or other legally protected status. For individuals with disabilities who would like to request an accommodation, please visit