Writing · X / Twitter · Email
I build real-time and distributed systems where latency, correctness, and product taste all matter.
My strongest interests are fraud prevention, distributed systems, search, real-time systems, and practical AI. I like backend-heavy projects that look like small production systems: ad-serving infrastructure, replayable crypto-market pipelines, event streams, freshness checks, ranking/search workflows, and useful interfaces on top of live data.
I write mostly Java, Go, Python, and TypeScript. Java is the language I have used the most day to day; my current public projects lean Go and Python because they fit the systems I am building right now. I am not trying to collect languages; I use them around one theme: turning messy real-world signals into systems people can trust.
| Project | What it shows | Stack |
|---|---|---|
| bidflock | An ads infrastructure project: OpenRTB-shaped bid requests, service boundaries, Redis hot-path reads, Kafka/Redpanda events, ClickHouse analytics, and latency-oriented auction design. | Go, gRPC, Redis, Redpanda/Kafka, MongoDB, ClickHouse |
| ticksense | A replayable crypto-market lakehouse for market analytics, freshness monitoring, and reproducible data workflows. | Python, crypto market data, analytics |
| portfolio-website | My personal site and public surface for projects, writing, and experiments. | Python, web |
- Systems that can be explained. I like codebases with clear boundaries, visible data flow, and failure modes you can reason about.
- Live data with replay. Dashboards are more useful when the underlying pipeline can be audited, rebuilt, and tested against history.
- AI as leverage, not wallpaper. I am interested in AI features that help people inspect, decide, and ship faster.
- Small projects with real constraints. Latency budgets, backpressure, freshness, observability, and correctness are more interesting to me than toy demos.
I am building toward a profile centered on real-time, data-intensive systems:
- ads infrastructure and budget-safe bidding paths
- crypto-market data infrastructure
- fraud-prevention signals and decisioning
- search, ranking, and retrieval workflows
- event-driven analytics and lakehouse-style workflows
- practical AI interfaces for understanding fast-moving data
I use writing to clarify what I am learning and why I am building things. You can find longer notes on Medium.
If you are working on systems, data infrastructure, applied AI, or products that sit close to live data, I would be happy to talk.
