I have 12 years experience in applied ML/AI (mostly at DeepMind and Google). I have led significant advancements in consumer and enterprise products – over 20 launches to YouTube, Google Shopping and Pixel phones.
My strategic foresight in early adopting LLMs and NLP advancements culminated in pioneering on-device GenAI features. I built and led senior ML teams in US and Asia.
My leadership extends across a vast spectrum of activities, from orchestrating ML strategy and driving innovation through hackathons to leading high-stakes projects and fostering a culture of continuous learning.
My commitment to development is evidenced by extensive coaching and spearheading the largest - more than 500 professionals - ML community in APAC at Google.
Collaborated with multiple teams to introduce AI and deliver impact into their products for the first time. Additionally, started an AI incubator, running 6-week sprints to turn one-liner ideas into robust product concepts. Managed up to 10 sprints per quarter, guiding teams through ideation, user research (UXR), and prototyping.
My international experience and cultural adaptability, combined with a growth mindset shaped by tackling high-risk projects, make me a resilient leader, in the rapidly evolving AI landscape.
Before moving into the nascent AI industry in 2013 and experiencing its spectacular rise for the past decade, I worked at fintech building applications for credit derivatives team at Bank of America Merrill Lynch.
Recorder Summarization – the first foundational language models on mobile devices (press), awarded one of the 10 tech innovation at Google in 2023. LLM-based weather summarization and diffusion-based weather background generation. Bandit-based recommender for Pixel.
NLP-based bug management tools lowered fix latency and costs. LLM-based root cause analysis tools. LLM-based test scheduling offered a double-digit cost reduction. LLM-based test code generation.
Bandit algorithms for YouTube ad timing brought in the biggest CTR increase in product history. Multiple launches in YouTube Ads Safety and YouTube Trust & Safety contributing $xxxM. Noisy labels techniques leading to a publication: arxiv
Neural-network based recommendations – 5 launches, each with a single-digit increase to #queries, #clicks, #revenue, #interactions. Bandit algorithms for whole-page layout optimization – a launch resulted in decrease in abandoned queries.