Jan 2023 – Present · 3 yrs 3 mos · Seattle, Washington, United States · On-site

• Leading core fabrics R&D of Amazon Rufus Gen AI.

Jan 2023 – Jan 2023 · 0 mo · Singapore · Remote

Jul 2017 – Nov 2020 · 3 yrs 4 mos · Mountain View, California, United States · On-site

• Working on Tensorflow/TPU Kernels and making TPU available in GCP as TPU VM.

Jan 2017 – Feb 2019 · 2 yrs 1 mo

• Helped the baby going from 0, as a solo founder for the first 12 months, raised the first $1M and $5.3M round of funding from world class VC and closed half millions dollars contract in the first 12 months deploying AI in real-world to some of the Billion dollars industry in South East Asia. On indefinite leave and passing the torch over and taking care of my health since early 2019 right after raising $5.3M round. “Health is something you take it for granted until it isn’t”.

Jun 2016 – Jun 2023 · 7 yrs · Mountain View, California

• Working on the Large Language Model (LLM) training and serving infrastructure.
• Engineering leader with hands on experience on technologies such as TPU, Cloud ML, Tensorflow and PyTorch runtime, and a proven track records delivering impacts through name brand products such as Youtube, Ads, GCP Cloud ML (Video, NLP, Document AI), Tensorflow.

Mar 2014 – Present · 12 yrs 1 mo

• As an entrepreneur, incubating and investing in AI technologies, working on Stealth AI startups.

Jan 2014 – Jun 2016 · 2 yrs 5 mos

• Deep Learning and Distributed Machine Learning project.
• How often do you get the opportunity to be part of developing the next big thing in computing? And how often can you join a dynamic team working on bleeding edge technology with a track record delivering? Do you dream scalability, consistency and availability? We are looking for highly experienced Distributed Systems design and coding gurus who are not afraid of pushing the state of the art in the field. You will work side by side as a team with other strong developers to design, prototype, and code the next greatest system. The ideal candidate will need to be able to work end to end, up and down the stack on focused high profile projects, be self-sufficient but highly accountable. Come join us to enjoy the excitement of an environment where your work can impact hundreds of millions of users and potentially change the world!
• We are focused on building scalable infrastructure for large-scale machine learning. We build real distributed systems, not research prototypes. We are looking for candidates with 5+ years of experience designing, coding, and shipping professional software or equivalent experience. Solid knowledge of distributed systems concepts, OS internals, proficiency in C/C++, a BA/BS or MS degree in Computer Science or Engineering are required. Proven expertise with networking, API design, scalable services development, scalable systems is ideal. Familiarity with machine learning basics is desirable but not required. But both the willingness to learn and the ability to learn quickly are essential. Must be comfortable working as a member of a small, high intensity team without traditional PM and test support.
• Group publications:
• https://www.usenix.org/conference/osdi14/technical-sessions/presentation/chilimbi
• http://www.wired.com/2014/07/microsoft-adam/
• http://research.microsoft.com/en-us/news/features/dnnvision-071414.aspx

Jan 2013 – Jan 2014 · 1 yr · Seattle/Cuppertino

• I am working on an exciting area crunching and converting BigData into Marketing $Insights$!
• BlueKai is the world's first complete enterprise BigData for marketing.
• BlueKai offers its customers a solution for managing and activating all their 1st and 3rd party data for use in their marketing and customer interactions. BlueKai represents the only end-to-end SaaS solution for marketers and publishers who are looking to maximize their cross-channel marketing efforts and create a proprietary solution for unlocking reach, scale and efficiency in their data.
• BlueKai has led the data-driven marketing category since 2008 when it launched the world's first data exchange and then branched out to create the first data management platform (DMP) for marketers. BlueKai is currently trusted by many of the Fortune 30 corporations and thousands of other brands to activate their data and by hundreds of other companies for data management needs.
• BlueKai was most recently named to Forbes' 50 Most Promising US Companies and Fast Company's 10 Most Innovative Global Companies in Big Data. In addition, BlueKai has been named to the Always On lists for 3 consecutive years. BlueKai is also named #12 on the @Deloitte #Fast500 - fastest growing companies list North America - #4 in Software.
• Specialties
• advertising, big data, marketing, DMP, data, insights, analytics, campaign optimization
• Website
• http://www.bluekai.com

Jan 2006 – Jan 2013 · 7 yrs

• 2012-2013 - SQL/Windows Azure Storage
• Technical role in Table layer, maintaining and optimizing the index management. This area deals with traffic load balancing, performance, index maintenance, capacity, distribution and monitoring.
• Technical lead on a short-term start-up project to develop a rich interactive box-line workflow editor web app (HTML5/SVG/Typescript/jQuery/KnockoutJS) for Machine Learning/Analytics cloud service.
• 2008 - 2012 - Bing IndexServe
• One of the key members of the BigIndex project Maguro: a system for indexing and searching over very large text collections (http://dl.acm.org/citation.cfm?id=2433486). Working on core pieces of the technology stack (Matching Engine, Network Subsystem, and Query Execution Engine). One of the early contributors on Architectural and Implementation Design, E2E System Integration, Perf optimization, Experimentation, Monitoring & UI/Tools development.
• Large-scale distributed index distribution & management system (involving data transfer ~peta bytes across ten of thousands of live servers)
• Owning critical portion of Bing query execution pipeline (query parsing, compilation, language, and grammar).
• Publication of Bing advanced query operator syntax documentation (http://msdn.microsoft.com/en-us/library/ff795620.aspx)
• Hands on experience with development/modification of critical parts of search engine index serving stack
• 2006-2008 - Messenger Server
• Large-scale service monitoring & deployment infrastructure
• Data visualization control development
• Geo-distributed PubSub service
• Live-site system with several hundreds users world-wide

Jun 2004 – Dec 2005 · 1 yr 6 mos

• Helping the company raising capital; developed the prototype of SpeedAlert system.