Oct 2018 – Sep 2019 · 11 mos · Menlo Park, California

• Product TPM at the Core ML team to bring broad impact for business integrity (Ads, Community, Pages, Feeds, Marketplaces) across Facebook's family of apps: Facebook, Messenger, Instagram, and WhatsApp, in order to reduce societal, political, personal, sentiment, and financial harms.

Jul 2014 – Oct 2018 · 4 yrs 3 mos · San Francisco Bay Area

• Director 2017 ~ present | Sr Manager 2014-2016
• Strategy and Technical Product Management roles in the R&D AI Team, responsible for setting the strategy, securing funding, recruiting the product team, and defining key product verticals to pursue. Led and delivered AI applications in manufacturing, digital marketing, and retail. Created and leading the product execution of SBrain, a data-scientists friendly collaborative AI platform to easily build and deploy AI models at scale.
• Product & Strategy Leadership
• Set the vision and strategy of the AI Team. Built buy-in internally from the team and from CEO, COO, SVP, VP of R&D
• Grew the AI Team’s technical and product influence to become regarded as “cradle of innovation” by CEO
• Created and leading Samsung’s SBrain initiative for building and deploying Perception and Decision models at scale
• o Led customer development, project de-risking, and alignment towards a product-market-strategy fit
• o Prioritized user stories and features development using lean product management
• Led business development with external & internal clients, and led the interactions with academia
• o Drafted SOWs, interfaced with Legal, negotiated contracts & terms, and performed M&A due-diligence
• o Led interactions with academic advisors with Professors from Berkeley/OpenAI, and Harvard
• Technical Product Management Leadership
• Led applying Reinforcement Learning (TRPO) to optimize advertising spend for online marketplaces. Patent filed
• Led applying Reinf. Learning (A2C) to energy-workload operation in storage servers, paper published at ECML-PKDD’18
• Led 1st introduction of Contextual Multi-Armed Bandit and Deep Learning for offline Ads in Korea (’15)
• Tech stack: TensorFlow, Rllab, Python, Scala, Spark, Docker, Kubernetes
• Keynote speaker at the AI World Forum Disrupt ’18, and AI & Digital Innovation ’18 on “Reinforcement Learning in the Wild and Lessons Learned”

Mar 2008 – Jul 2014 · 6 yrs 4 mos · San Francisco Bay Area

• Principal Solutions Architect ’14 | Sr. Staff Systems Engineer ‘13 | Staff Systems ‘10-‘12 | Sr. Systems Engineer ‘08-‘10
• I developed algorithms and production software to diagnose and optimize full broadband PHY-layer access network for telecommunication service providers. Founded by my Stanford PhD co-advisor, I joined ASSIA when it was pre-product as one of its founding core R&D team, and I helped it grow to manage 70M broadband lines when I left (12% of the world’s fixed broadband). It is now managing 100M broadband lines.
• Designed and architected multi-million dollar solutions for customers in North and Latin Americas, Europe, MENA, China, and South East Asia. Led RFIs and RFPs responses to major global telcos
• Expert in full algorithms development cycle: from data collection to algorithm design, implementation, integration, testing, introduction in mission-critical production environment, and evaluation. Stack: MATLAB, Java, SQL, BIRT
• Due to the startup nature, I worked on versatile projects from algorithms engineering, to software engineering, DB backend, features and scalability testing, customer deployment, and trials evaluations and analysis
• Piloted a machine learning based churn predictor in the field to reduce churn at a major US telco
• Mined Twitter data to gain geo-temporal insights of broadband service quality (used Gnip, later acquired by Twitter)
• Enabled data to tell a story and a voice to drive decisions, and opportunities.

Jul 2006 – Sep 2006 · 2 mos · Santa Clara, CA

• Intern at the Distributed Communication group. I researched and implemented MATLAB algorithms for multi-user and multi-hop communication. Paper published at the Conference on Information Sciences and Systems.

Jul 2005 – Sep 2005 · 2 mos

• Intern reporting to the CTO. I joined at the early stages of the company. I conducted technical due-diligence on MIMO algorithms and the system design. Beceem developed WiMAX & LTE chips, later it got acquired by Broadcom for $316M.