Founder of Token Canopy. I build infrastructure for AI agents and the people who work with them.
- Token Canopy — a collaboration layer for people and AI agents.
- e2a — an open-source email gateway for AI agents. Explore e2a →
- AgentDrive — a persistent workspace where agents can store, find, and share artifacts.
At Google (2020–2026), I was a Senior Software Engineer working across cloud automation, data infrastructure, and AI-assisted data engineering—including Google Cloud Workflows. My data-infrastructure work focused on Google's core data-processing platform, now part of its AI foundation, which powers some of Google's largest-scale data pipelines. I also initiated and led the data engineering agent project for this platform, taking it from a hackathon pitch to a fully funded team.
In 2026, I briefly joined Resolve AI before choosing the founder path and starting Token Canopy.
I also contributed to platform development for Stanford ADRC's SPHERE Research Data Hub, a research platform for exploring privacy-preserving synthetic, multi-omics data.
- Google-scale compute and storage: Hands-on experience with Borg, Colossus, in-memory file systems, and high-throughput storage infrastructure, including file sharding, lifecycle management, and compaction.
- Databases and query systems: Experience with F1, Napa, Spanner, BigQuery, DuckDB, PostgreSQL, MySQL, Supabase, and DynamoDB—spanning distributed SQL, analytical storage, real-time ingestion, and relational and key-value workloads.
- Distributed systems: Familiar with Raft and Paxos consensus, distributed hash tables, consistent hashing, replication, partitioning, and failure recovery. Implemented Chord, a scalable distributed lookup protocol for mapping keys to nodes.
- Tracing and observability: Built distributed tracing infrastructure for large-scale data-processing engines, covering execution-level and event-level telemetry. Scaled ingestion with caller-side caching and buffering, transactional message-queue batching, file sharding and compaction, and engine-level filter pushdown.
- Cloud infrastructure: Built and operated systems using Cloud Run services and jobs, Cloud Tasks, Cloud Scheduler, Cloud Workflows, Cloud Storage, Cloud SQL, Artifact Registry, Secret Manager, Terraform, Docker, GitHub Actions, Cloudflare, Prometheus, Grafana, and OpenTelemetry, plus AWS Lambda, API Gateway, S3, DynamoDB, and SES.
- Princeton University — M.S.E. in Computer Science
- University of Washington — B.S. in Electrical Engineering
Languages
Cloud & platform
Infrastructure & delivery
Data & storage
Observability
