SYSTEM / salil-hiremath
ROLE / computer science student · applied AI builder
FOCUS / agentic systems · SRE automation · GraphRAG · ML
MODE / turn messy signals into dependable software
I build practical AI systems that do more than chat: they inspect context, choose a next action, use tools responsibly, and leave an audit trail. My current frontier is the intersection of autonomous agents, reliability engineering, retrieval, and applied ML.
| Signal | What it means |
|---|---|
observe |
Telemetry, documents, repositories, and real-world inputs become usable context. |
reason |
Agents decompose ambiguity into decisions, checks, and bounded actions. |
ship |
A useful outcome includes validation, a handoff, and a path to production. |
Open the builder log
- 🎓 B.Tech CSE student in India, exploring the space between intelligent systems and dependable engineering.
- 🤖 Working on multi-agent SRE workflows, GraphRAG, GenAI assistants, and MCP-powered developer tools.
- 🧭 Interested in products where AI has a real job to do—not just a text box to fill.
- 📬 Best way to reach me: [email protected].
01 — Always-On-Call SRE · autonomous reliability engineering
An agentic SRE workflow that moves from anomaly detection and root-cause analysis to validated fixes, merge-request handoff, and postmortems. Built around the idea that production incidents deserve an evidence trail—not an opaque answer.
02 — Workforce MCP · specialist context on demand
An MCP server that gives coding agents access to focused specialist packs, collaborative pods, and workflow context—so the right expertise is available at the point of work.
03 — Data & retrieval experiments · GraphRAG, search, and applied ML
I also explore contextual search, document intelligence, voice agents, computer vision, and IoT/ML applications. Browse the repositories for the latest experiments and iterations.
Open to thoughtful conversations about agent systems, ML products, and engineering that earns trust.