Software Engineer
Robinhood
Nov 2024 - Present Toronto, ON
- Built the centralized ML/Agent governance platform supporting the company's RAG applications. Designed model registries and retrieval-aware schema validation for LLM agents.
- Standardized DB transaction patterns and resolved sequence overflow issues in high-scale services; delivered a centralized DB package with SLO alerts, resulting in zero on-call tickets for DB operations since rollout.
- Implemented a cross-system Rules Engine and validation APIs (Go/Protobuf) to encode complex legal policy decisions, replacing manual reviews with deterministic backend logic.
- Integrated LLM providers (OpenAI, Anthropic, Vertex) and vector/retrieval toolchains. Operationalized Braintrust ETL for continuous evaluation of hallucination, bias, and jailbreaks.
- Authored a protobuf-based codegen framework (Go/Python) to centralize enums/tag maps, guaranteeing cross-language synchronization and eliminating schema drift across distributed services.
- Led comprehensive refactors to raise deployment safety, increasing unit test coverage from ~10% to ~70% in critical backend paths.
Go Python Protobuf RAG LLM PostgreSQL Kubernetes