Remote: Yes (open to hybrid or remote)

Willing to relocate: Yes (SF, NYC, Austin, or other US tech hubs)

Technologies: Python, JAVA, C++, FastAPI, TypeScript, React, PostgreSQL, SQLite, Docker, AWS, PyTorch, LLM APIs (OpenAI/Anthropic), LangChain

Résumé/CV: https://darshannere.com | https://github.com/darshannere

Email: darshan.nere2@gmail.com

I build AI agent infrastructure. Most recently: Tracea (tracea.dev), an open-source observability platform for AI agent teams that patches the HTTP transport layer to trace LLM calls without any framework changes — no SDK wrappers, no code changes, just install and instrument. Also built ObservAgent (1000+ active users, open source).

Research side: co-authored a paper on failure-aware penetration testing - training LLMs to emit typed failure tokens so failure recognition becomes a structured recovery signal rather than an implicit state.

Prior: full-stack engineer shipping software to 50k+ users.

Looking for full-stack or AI/ML engineering roles at early-stage or growth-stage companies working on production AI systems - LLM integration, agent infrastructure, or observability. High ownership, real users, hard problems.

Remote: Yes (open to hybrid or remote)

Willing to relocate: Yes (SF, NYC, Austin, or other US locations)

Technologies: Python, FastAPI, TypeScript, React, PostgreSQL, SQLite, Docker, AWS, PyTorch, LLM APIs (OpenAI/Anthropic), LangChain

Résumé/CV: https://darshannere.com | https://github.com/darshannere

Email: darshan.nere2@gmail.com

I build AI agent infrastructure, most recently Tracea (tracea.dev), an open-source observability platform for AI agent teams that patches the HTTP transport layer to trace LLM calls without framework changes. Also built ObservAgent (300+ users, open source) and published research on GRPO fine-tuning for medical reasoning (+11% accuracy over SFT baseline).

Looking for full-stack or AI/ML engineering roles at early-stage startups or growth-stage companies where I can work on production AI systems with high ownership and real impact. Strong preference for roles involving LLM integration, agent infrastructure, or AI observability.

Caude writes JSONL transcripts but they're not usable in real time.

So I built ObservAgent. It's a local dashboard that gives you live visibility into every Claude Code session without touching your code.

How it works:

Claude Code has a hooks system (PreToolUse, PostToolUse, SubagentStart, SubagentStop). ObservAgent installs a tiny Python relay that fires a fire-and-forget POST to a local Fastify server on every hook event. The server stores events in SQLite and streams them to a React dashboard over SSE.

npm install -g @darshannere/observagent observagent init # installs hooks into ~/.claude/settings.json observagent start # opens dashboard at localhost:4999

That's it. Run any Claude Code session and watch it live.

What you get:

- Real-time tool call log with latency per call - Cost tracking — token usage per model (including cache read/write) auto-calculated against current Claude pricing - Agent tree — session + subagent hierarchy with per-agent cost rollups - Session history grouped by repository with one-click replay and JSONL/CSV export - Health panel with p50/p95 latency and error rates

Everything binds to 127.0.0.1. No cloud, no telemetry, no API keys needed.

GitHub:https://github.com/darshannere/observagent npm: npm install -g @darshannere/observagent

Happy to answer questions about the hooks architecture or how I handle cost attribution across subagents (trickier than it sounds).

Comments URL: https://news.ycombinator.com/item?id=47391414

Points: 2

# Comments: 0