Explore my work · Try Incident Lens · Let's connect
I build tools that help engineers ship with confidence: reliable distributed workflows, reusable platform patterns, and AI tooling with verification built in.
I'm Vijay, a hands-on software engineer in Richmond, Virginia, with 15+ years across full-stack development, platform modernization, and production operations. My work spans C#/.NET, TypeScript, React, Node.js, Python, AWS, and Azure.
| Project | The problem it tackles | Take a look |
|---|---|---|
| Incident Lens | Production failure patterns hide in ordinary code. A VS Code extension uses Claude to flag risks such as unbounded retries, missing idempotency, and N+1 queries. | Source & setup |
| Reliability Workflow Lab | Reliability concepts are easier to understand when you can see retries, rate limits, and idempotency in action. | Open the interactive lab |
| TSA Wait Time Intelligence | A portfolio dashboard brings together wait-time prediction, retrieval, and an AI analyst interface. | Explore the dashboard |
| TypeScript Backend Lab | Small simulations make queue backpressure, idempotency, and retry/DLQ behavior concrete. | Try the backend lab |
Useful AI needs a delivery system around it. I built and operate Mission Control, an internal agent-orchestration platform that takes work through planning, implementation, verification, and engineer review. Its baseline-relative verification gate distinguishes existing repository failures from regressions introduced by a change. Internal project; source is not public.
Reliability is a product feature. My day-to-day work includes asynchronous workflows, observability, incident response, and reusable patterns that make the next team's job easier. Incident Lens turns that production perspective into an open-source developer tool.
Context should travel with the work. I care about clear design decisions, small reviewable changes, codebase knowledge, and tools that help engineers contribute to unfamiliar systems.
I'm growing the public side of that work through practical examples, regression tests, clearer documentation, and small product improvements. The current focus is Incident Lens: making reliability feedback easier to understand and act on.
Engineering note: Cancellation hygiene: three small checks that matter — a runnable, offline example explaining the failure cases behind my LangGraph.js SDK contribution.
If you're building developer tools, AI platforms, or systems that need to stay dependable under load, I'd enjoy comparing notes.