San Francisco, California, United States
Visionary AI Product Leader | Bridging Frontier Research to Global Consumer Products I am an AI Product Leader and Ph.D. specializing in the 0->1 incubation of generative AI, agentic platforms, and multimodal systems. With a unique blend of deep technical expertise and executive product strategy, I bridge the gap between foundational AI research and scalable, market-defining products. My career is defined by driving Research-to-Product (R2P) pipelines, optimizing massive compute investments, and delivering global-level impact at some of the world’s leading technology companies. Global Impact & Scale at Meta and Apple At Meta’s Superintelligence Labs (FAIR), I lead the AI strategy and multi-year roadmaps for frontier models, world models, and agentic systems. Partnering directly with VP-level leadership, I orchestrate the optimization of 20k+ GPUs and $30M+ in OpEx. Previously at Apple, I defined the 5-year AI platform vision for the App Store, driving the deployment of server and on-device ML models that optimized acquisition, retention, and monetization for over 1.4 billion users while spearheading global privacy and compliance standards. Technical Innovation & Research My technical foundation is rooted in a Ph.D. in AI and Computational Intelligence. At Meta, I co-authored strategic research with Yann LeCun and drove foundational R2P work in VL-JEPA, Computer Use Agents, and Action100M. As an Adjunct Professor and Director of the AI Minor Capstone at USC, I instruct the next generation of engineers in advanced Technical Generative AI, LLMOps, and SOTA architectures, and GenAI products, guiding students through the 0->1 lifecycle of building GenAI products. Organizational Leadership I thrive in extreme ambiguity and highly matrixed organizations. I have a proven track record of scaling global teams from the ground up, translating complex technical trade-offs into data-backed capital allocation, and aligning cross-functional engineering, research, and legal teams to deliver robust AI ecosystems. Core Technical & Leadership Competencies: • Frontier AI & Architectures: Generative AI, Agentic Workflows, World Models, Multimodal VLMs, Transformers, RAG, RLHF. • AI Product Strategy: 0->1 Incubation, Research-to-Product (R2P), Compute Investment Allocation, LLMOps, A/B Testing, Data Flywheels. • Organizational Leadership: Executive (VP+) Alignment, Matrix Management, Global Team Building, AI Governance & Privacy (GDPR/PIPL). • Applied Domains: Predictive Analytics, NLP/Speech Processing, Recommendation Systems, Subscription Growth, Risk Assessment.
Lead of the largest research org at FAIR with 12 teams. Managing AI Research, Product Management, and Program Management end-to-end in partnership with other MSL orgs, TBD Lab, and Wearables orgs. Focus research areas in addition to pillar-level/leadership efforts for Frontier Models, Wearables and exploratory research: Self-Organizing Agents, Computer Use, World Models, Memory, Personalization, White-Box Distillation, Proactive AI Assistant, Published 6 papers listed below including with Yan Lecun! Product Management: - Develop multi-year strategy and roadmaps for research orgs contributing to frontier models and Meta Flagship AI glasses; lead X-org VP handshakes & sponsorships. - Own and drive R2Ps, building internal products, evaluation envs, benchmarks, Reinforcement Learning envs, training datasets, research frameworks, data annotations, and DPO. Leadership: - Lead VP reviews to secure VPs’ buy-in to sponsor research pillar vision & projects, partnerships with other orgs and flagship R2Ps, managed escalations with deputy C-suite. - Led a part of building Meta’s frontier models on the roadmap for public launch! - Manage teams of Machine Learning Engineers for AI coding research, Software Engineers, Engineering Managers, Compliance and Privacy PMs, and Research/Data Specialists for special projects. AI Research and Individual Contributions: - Building state-of-the-art personalized AI capabilities for agents, including shipping artifacts to Muse Spark and future models. - Initiated and organized an advanced AI research learning series for the FAIR Foundation org, covering hands-on coding from cluster access and infra to running fine-tuning jobs and evals on Nvidia GPUs. - Created/vibe coded and deployed internal data visualization products and dashboards with Claude Code for the entire FAIR organization, serving hundreds of researchers on a daily basis. - Coauthored research proposals and plans with research directors - Published 6 papers listed here
- School of Engineering I spearheaded an advanced technical course focused on state-of-the-art Generative AI and NLP techniques as well as building AI products end-to-end. Students obtained the following knowledge and hands-on experiences: -Training, Evaluation, and Fine-Tuning of Large Language Models (LLMs): Developed deep expertise in training advanced language models like ChatGPT and BERT, utilizing transformer architectures and employing tools such as Hugging Face for efficient model management. Focused on detailed training steps, including data augmentation, the use of Reinforcement Learning from Human Feedback (RLHF), Proximal Policy Optimization, and comprehensive evaluation metrics. -Advanced Prompt Engineering, Ethical AI, and Model Safety: Achieved proficiency in advanced prompt engineering techniques such as few-shot learning and chain-of-thought reasoning. Developing robust ethical AI frameworks, focusing on model safety to prevent data leakage and prompt injection, ensuring responsible and secure AI deployment. -Retrieval Augmented Generation (RAG), Conversational AI, and Semantic Technologies: Implemented RAG to elevate the performance of conversational AI systems, integrating it with cutting-edge speech recognition and semantic analysis for enhancing NLU. Utilized LangChain and Llama for customizing language models, thus pushing the boundaries of question answering and conversational capabilities. -Technological Integration and Scalable NLP Solutions: Acquired hands-on experience deploying NLP models through APIs, using tools like MongoDB, ngrok, and Postman for robust API management. Introduced LoRA and QLoRA for model quantization, facilitating the deployment of large models at scale. -Applications: Information retrieval and search, Sentiment analysis, Speech processing, Conversational AI, Text classification, Question answering, Machine translation, Text summarization, Prompt Engineering, Business applications, Vector Database, APIs
https://www.crystalytic.ai/ Industries served: Healthcare, Hospitality, E-Commerce, Entertainment, Manufacturing, Software Services, Consumer Electronics • AI and Gen AI products: Hired and managed ML engineers and UX/UI designers to build AI solutions and products for clients to enhance/predict sales and optimize inventory resulting in 5-15% increase in sales and 10-23% reduction in product waste in restaurants industry. • A-Z Management: Led defining business objectives and MVP, priorotized features based on user stories and managed design and engineering to deliver high-quality, well-designed products utilizing emerging AI and Gen AI technologies that meet customer needs and aligned with analytics insights. Experience: Sales Prediction & Optimization, Marketing, Risk Analysis, Recommendation Systems, Motion Analysis & Forecasting, User Acquisition & Engagement, Medical Diagnosis
Growth marketing - Machine Learning: content personalization, recommendation and financial models for worldwide App Store monetization and engagement. Acquisition and engagement models for subscription services - Focus: Managed strategizing, building, and deploying App Store and Apple Arcade user-facing Machine Learning and data models in user journeys end-to-end in accordance with privacy guidelines, GDPR, PIPL, and local regulations around the globe
Reported to Executive VP for critical revenue generating clients, owned and drove seven strategy roadmaps for healthcare digital products, supervised 15+ employees in US and India, managed budget of $X.XM/yr, and successfully delivered end-to-end from ideation to customer delivery. Increased follow-on project by $X.XM by effectively managing existing customer relationship, translated data into meaningful insights for clients and stakeholders and articulated complex structures and ideas clearly to court continued collaboration. Built 0-1 frameworks, defined product vision, wrote specs, integrated customer’s feedback, and scaled by providing leadership and training, resulting in enhanced customer satisfaction, shorter delivery time of products and 35% cost saving.
Drove average of four project roadmaps and backlogs annually, documenting and managing requirements and client stories to ensure prioritized delivery of business value and delivering projects up to $X.XM. Created ML regression models, performed hypothesis testing, while honing analytical skills and systems thinking to assist healthcare clients with identifying hazardous locations concerning applicable regulations for upgrade purposes. Managed all social events of the company and organized two holidays parties in JW Marriott DTLA and Hilton Pasadena, leading a team of 10+ and preparing all the contracts for vendors