Lead AI Engineer · Los Angeles
Building production AI systems — from on-device ML to cloud-scale data infrastructure. Currently focused on RAG architectures, real-time video processing, and the end-to-end tooling that makes AI products actually ship.
Consumer AI Products
iOS apps with on-device ML (CoreML, Vision) — real-time video processing, intelligent editing, and the kind of fluid UX that makes AI feel invisible.
Retrieval-Augmented Generation
Production RAG pipelines for complex document analysis — vector search, citation traceability, semantic chunking that works on messy real-world data.
Event-Driven Data Platforms
Dagster-orchestrated lakehouses on Databricks. High-velocity ingestion from social APIs. The unsexy infrastructure that makes analytics possible.
Developer Tools
CLI-native monitoring, voice-to-code interfaces, and the small sharp tools that make engineering less painful.
| Domain | Stack | Notes |
|---|---|---|
| AI Video Platform | Swift · CoreML · Vision · AWS | Full-stack consumer app — on-device ML for real-time video analysis, serverless processing pipeline |
| Legal Discovery RAG | Python · LangChain · AWS | Document analysis with citation-level traceability |
| Social Analytics Lakehouse | Databricks · Dagster · Spark | Medallion architecture for real-time social metrics |
| Infrastructure Health TUI | Go · Bubble Tea | AWS monitoring for the terminal |
| Real-Time Social Ingestion | Python · Lambda · EventBridge | Sub-minute data capture from Meta APIs — the pipes that feed the lakehouse |
Most repositories are private.
Languages: Python, Swift, Go, SQL, HCL
ML/AI: CoreML, Vision, LangChain, PyTorch
Infrastructure: AWS (Lambda, Step Functions, ECS), Databricks, Terraform
Data: Dagster, dbt, Spark
Learning: Rust, more Go, whatever's next
Open to interesting problems.