I build production backends and AI systems for teams that need to ship — not demos that look good in tabs.
CS @ KIIT '27 · Currently at CWI Studios · Bokaro / Bhubaneswar.
| What it is | Why it matters | Stack | Live | |
|---|---|---|---|---|
| 🏭 IndustryERP | Production ERP at a steel fabrication business | 7+ employees use it daily. 172 endpoints. Replaced their spreadsheets. | FastAPI · PostgreSQL · Cloud Run | ↗ |
| 🛡 OneAlert | OT/ICS cybersecurity SaaS — AI remediation, SBOM, compliance | 26⭐ · 166 tests · 10 shipped phases · MIT licensed | FastAPI · React · Cloud Run | ↗ |
| 🤖 TaskForge | Multi-agent crisis logistics system | Built for Google Cloud Gen AI Academy APAC 2026 (Cohort 1) | Gemini 2.5 · MCP · pgvector | ↗ |
| 📞 Redline AI | AI emergency dispatch IVR (fork) | Contributed Gemini integration + CI/CD + production deployment | FastAPI · Twilio · Whisper | ↗ |
| 🚁 SkyDash | 20Hz MAVLink drone telemetry dashboard | Real-time streaming, backpressure handling | FastAPI · React · WebSockets | ↗ |
- Reported and specified a cross-tenant RBAC isolation regression suite for AiSOC: proposed endpoint-level tenant-isolation tests, BYOK/alerts coverage, and a nightly security CI gate in #159, later closed by merged PR #197.
I treat the shop floor and the customer call as primary sources, not the GitHub trending page. Before writing the ERP, I spent weeks interviewing operators at a steel plant. That's why the system actually gets used — and why I reach for MCP, agentic AI, or real-time streaming only when they solve a real problem.
Python · FastAPI · PostgreSQL · Redis · Docker · GCP Cloud Run · Cloud Build · GitHub Actions · Prometheus · Grafana
Comfortable with MCP, agentic systems, RAG, pgvector, and real-time streaming pipelines when the problem calls for them. RHEL-certified. GitHub Developer Program member.
🛠 Building Blue — internal platform powering CWI Studios' operations 🌱 Open to full stack / systems / AI engineering internships where I can ship to real users from day one