David Cheng Email & Phone Number

Bellevue, Wa, Us

Devens, Massachusetts, Us

• Lead architectural design for core business systems, automating and streamlining complex processes.• Developed an in-house dynamic business configuration management platform.• Design and implemented a customer support agent using generative AI technologies.

Seattle, Wa, Us

• Led 2 engineering teams to deliver ML orchestration solutions for continuous training, parallel computing and automated model building in Alexa. Delivered unified tool for infrastructure, permission and artifact management used by 1400+ modelers in Alexa.• Analyzed architecture of legacy experimentation framework, led team of developers to increase stability of framework to 99%. Led team of 9 to re-architect the experimentation framework, leveraging open source Dask library, AWS SageMaker and AWS ECS. • Coordinated with engineering teams in Alexa and AWS to drive technology vision on enterprise-scale MLOps solutions.

Seattle, Wa, Us

• Technical lead for data center automation systems. Developed tools for process orchestration and work management using Camunda BPM and Amazon web services.

Seattle, Wa, Us

Founding member of a new initiative in Alexa Machine Learning to make Alexa smarter and more natural to engage with. Worked with machine learning scientists to launch name-free skill interaction and CanFulfillIntentRequest SPI.https://developer.amazon.com/blogs/alexa/post/60e1f011-3236-4162-b0f6-509205d354ca/making-alexa-more-friction-freehttps://developer.amazon.com/blogs/alexa/post/4e6db03f-6048-4b62-ba4b-6544da9ac440/the-scalable-neural-architecture-behind-alexa-s-ability-to-arbitrate-skillshttps://developer.amazon.com/blogs/alexa/post/c870fd31-4f91-4a62-a137-8dfa70ca5e9c/hyprank-how-alexa-determines-what-skill-can-best-meet-a-customer-s-needhttps://developer.amazon.com/blogs/alexa/post/352e9834-0a98-4868-8d94-c2746b794ce9/improve-alexa-skill-discovery-and-name-free-use-of-your-skill-with-canfulfillintentrequest-beta

Seattle, Wa, Us

• Led team of 5 engineers to design and implement a self-service onboarding tool for OrderToCash platform, streamlining the onboarding process and reduce SDE effort involved in new onboardings • Led team of 4 engineers to design and implement validation workflows to replace existing manual processes. The workflows are used to verify migration of business event processing onto self-service APIs and ensure that the events are recorded in accounting sub-ledger correctly. Analyzed performance of validation logic and implemented optimizations to reduce latency by ~90 percent• Worked with ordering systems team to come up with a plan to enable incremental migration of shipment and refund events onto self-service APIs. Implemented matching and filtering logic in business event processing workflow to unblock the migration project• Designed and implemented system to analyze results from business event processing to discover accounting use cases and publish into AWS Cloudsearch to allow search and discovery of the use cases• Implemented improvements to Scrum process for team in order to provide more visibility and to prioritize on delivering value to business • Increased fault tolerance of remittance file processing workflow by migrating workflow to use Amazon S3 as file repository• Mentored interns to build diagnostic tools to provide visibility to the system and increase productivity of operations personnelTechnologies: Java, JAX-RS, SQL, Spring, Guice, Amazon web services (S3, SNS, SQS, Redshift, DynamoDB, Lambda, Kinesis Firehose, Cloudsearch), AspectJ, NoSQL, JSON, XML, Oracle, Hibernate, ETLM, JSP, JQuery, Ruby on Rails

Seattle, Wa, Us

• Lead developer in building workflows to enrich Kindle eBooks with additional navigation metadata. Coordinated with offshore teams to streamline manual creation/audit of navigation metadata and provide desktop/web tools for operations team. Implemented system to automate navigation metadata creation to reduce handling time of operations personnel• Led team of 5 engineers in building validation platform to analyze quality of Kindle eBooks. Designed a flexible data model to ease schema evolution and functionality to reconcile quality-related data from multiple sources• Coordinated with offshore teams to integrate operations workflow with the validation platform to record feedback from subject matter experts on spell check results, resulting in increased productivity for the subject matter experts• Volunteered to help unblock launch of pricing tool on Kindle self-publishing website by implementing book length estimation logic in validation platform • Designed and implemented system to automate mapping of physical book page numbers onto Kindle eBooks, reducing delay in page numbers generation from a week to minutes and increased glance view coverage of page numbers program by 10 percent• Enabled generation of physical book page numbers from ePUB page lists and worked with account manager to troubleshoot data quality issues • Led team in migrating existing systems to support next generation Kindle eBook format, proposed design to senior engineers to allow navigation metadata to be updated independent of book content• Implemented self-service diagnostic tools to increase productivity of operations personnelTechnologies: Java, SQL, Spring, Amazon web services (S3, SimpleDB, SNS, SQS, Simple Workflow, DynamoDB, Elastic MapReduce), AspectJ, NoSQL, JSON, XML, Oracle, ETLM, JSP, JQuery, Eclipse SWT

Seattle, Wa, Us

• Part of a team that launched Shopping-enabled Wikipedia pages(http://news.cnet.com/8301-13577_3-20024297-36.html). Designed and implemented component to relate content in Wikipedia articles to Amazon product pages.• Worked on backend services and data pipelines to support Artists pages, author pages and TV series pages on Amazon websites. Developed new widgets and modified existing ones to improve customer experience.• Migrated Amazon's career site search from a third party system to Amazon's internal search infrastructure.Technologies: Java, SQL, Spring, AspectJ, Amazon web services (S3, SimpleDB, SQS, CloudSearch), NoSQL, MySQL, Oracle, HTML, Perl

Seattle, Wa, Us

• Implemented new features on high throughput item management system that supports Amazon’s E-Commerce Platform and maintains Amazon’s massive product catalogs for all Amazon operated websites and its global fulfillment network. • Designed and implemented rules engine to reduce time it takes to add new business logic for reconciling product data from multiple data sources• Implemented self-service diagnostic tools for internal teams to troubleshoot issues, reducing operational burden of team membersTechnologies: Java, XML, XSL, SQL, Oracle

Richardson, Texas, Us

Implemented new features and improvements to service logic components in real time billing system on UNIX platform, using C/C++, Pro*C. Developed and maintained test tools, conducted performance tests and profiling experiments. Provided technical leadership and mentorship to new team members.

M. Eng, Computer Science

B.S., Computer Science

Education record