Launched Instagram Tune Your Algorithm
Led technical design and infrastructure for the US launch, taking the feature from inception through public testing to production in under six months.
Senior Software Engineer
Building AI systems that move from prototype to production.
I am a Fremont-based engineer focused on recommendation systems, LLM infrastructure, prompt optimization, content understanding, and scalable ML products.
Meta
Media Understanding Team
Google
Core-ML and Trust Systems
3.9
M.S. CS GPA at UMass Amherst
Selected Impact
Highlights from shipped systems, product launches, and production ML infrastructure.
Improved LLM prompt performance
Built prompt optimization systems used by 20+ product areas, delivering measurable gains through AI-driven tuning, rewriting, and reranking.
Reduced review latency
Owned AI and human hybrid review systems for abusive content in Search, reducing takedown time from days to minutes.
Experience
Recent roles centered on high-impact AI product launches, applied ML infrastructure, and production systems.
Media Understanding Team
- Led the technical design, infrastructure, and launch path for Instagram's Tune Your Algorithm experience.
- Scaled adoption to 1.1% of US Reels DAU within weeks of launch and helped drive a 30% week-over-week increase in daily TYA reel impressions.
- Developed the technical blueprint for Incremental Bio, targeting substantial GPU capacity savings.
Core-ML Team
- Designed LLM-powered prompt optimization systems across tuning, rewriting, reranking, multimodal inputs, and GenAI evaluation.
- Launched Prompt Iterative Developer Journeys and Prompt Execution APIs for automation pipelines and Search Review workflows.
- Served as a founding engineer for PromptLab, prototyping AI agents, tool invocation, reasoning-based prompt chaining, and reusable prompt galleries.
- Led production work across prompt execution storage, security, privacy, and compliance for enterprise LLM deployments.
Core Team, Sunnyvale
- Built A/B classifier evaluation infrastructure for model calibration and rollout across image and video moderation systems.
- Integrated PhotoDNA, fuzzy matching, and hash-based detection methods to improve detection scale and accuracy.
- Improved automation in human review workflows, reducing redundant review effort and saving operational resources.
Core Skills
A practical mix of applied ML, backend systems, product infrastructure, and evaluation workflows.
- LLM systems
- Prompt optimization
- AI agents
- Multimodal AI
- Recommendation systems
- ML infrastructure
- Model evaluation
- A/B experimentation
- Content safety
- Search review systems
- Production APIs
- Privacy and compliance
Education & Recognition
Academic foundation in computer science with repeated recognition for production impact.
University of Massachusetts Amherst
Master of Science in Computer Science. GPA: 3.9/4.0.
College of Engineering Guindy, Anna University
Bachelor of Engineering in Computer Science and Engineering. CGPA: 9.39/10.0.
Google Core Impact Award
Recognized with the team for actionable reporting work in child safety systems.
Peer and Spot Awards
Earned 25+ peer awards and spot cash awards across Google and Intuit.