• Working on the problem space for privacy-compliant data infrastructure supporting Meta Superintelligence Labs — defining scope, identifying gaps, and driving solutions across ingestion, enforcement, and lineage.

• Creating scope for E4 engineers by decomposing ambiguous privacy infrastructure challenges into structured workstreams, accelerating team delivery on MSL research timelines.

• Driving cross-functional technical alignment across Privacy Policy, Data Infra, and MSL Research — translating regulatory requirements into concrete infrastructure decisions.

Tech stack: C++, Hack, Python · Spark, Presto, Hive, Koski

• Built and owned bulk data ingestion tooling for Meta Superintelligence Labs (MSL) researchers, enabling privacy-compliant onboarding of training datasets at billion-row scale.

• Implemented data mitigation and filtering pipelines to detect and remove privacy-sensitive content from GenAI training data flows before model consumption.

• Contributed to enforcement implementation for Meta's Privacy-Aware Infrastructure (PAI), applying purpose-limitation controls across high-volume data pipelines.

• Worked with Privacy Policy and Data Infrastructure teams to ensure GenAI training workflows complied with GDPR, CCPA, and internal data governance standards.

Tech stack: C++, Hack, Python · Spark, Presto, Hive, Koski

• Worked on a multi-module Java orchestration platform managing application and infrastructure deployments for both internal teams and external Oracle Cloud customers

• Contributed to reliability and scalability improvements across the deployment pipeline, supporting complex multi-tenant cloud infrastructure at enterprise scale

• Operated in a fast-paced IC role spanning backend services, infrastructure automation, and cross-team coordination across Oracle Cloud's deployment ecosystem

Tech stack: Java, Python, Golang, Docker, Terraform, Maven

• ML engineer delivering homepage personalization systems for 100M+ users at 60K RPS, 3.6–4M active sessions, and 200ms SLA.

• Led for Immersive Hero — designed the ML algorithm from scratch driving homepage engagement; project was cited by CEO & President of Global Streaming JB Perrette in a press conference as a key personalization initiative.

• Spearheaded Top 10 Movies/Series — a content discovery feature that surfaced trending titles to users at scale, receiving coverage in Variety magazine at launch

• Built and owned Recommended For You and Because You Watched ranking systems using knowledge graphs and XGBoost, improving content relevance across the homepage feed.

• Developed and maintained homepage ranking infrastructure — the core personalization layer determining content order for every user session on the platform.

• Contributed to A/B testing and experimentation frameworks to validate model performance and drive data-driven decisions on personalization product direction.

• Designed, built, and shipped a full-stack distributed search application to production, used internally within AWS.

• Implemented a custom query parsing layer using ANTLR to handle complex search expressions, integrated with Elasticsearch for scalable indexing and retrieval.

• Built serverless backend infrastructure using AWS Lambda and API Gateway, enabling low-latency query execution at scale.

• Delivered end-to-end — from design through testing and production deployment — within a 10-week internship timeline.

Tech stack: Java, Python, AWS Lambda, API Gateway, Elasticsearch, ANTLR, Mockito