CS Student · Builder · ML Enthusiast
2nd-year Computer Science student at ANITS, Vizag. I build things that run — local AI desktops, LLM routers, real-time safety apps, medical routing engines, and multi-agent systems. Stack: Python, FastAPI, React, TypeScript, Kotlin + LangChain & LangGraph.
A privacy-first AI desktop assistant with a glassmorphic floating UI. Runs Qwen2.5 3B fully offline — no cloud, no data leaving your machine. OCR, voice control, and an ambient interface that layers over your entire workflow.
LLM Hub — route prompts through multiple language models with automatic fallback. Built-in auth, configurable model stacks, and a unified API over heterogeneous AI backends. One interface, every model.
Multi-perspective analysis agent — any situation or question is examined simultaneously through multiple domain-expert AI agents. Each agent owns a lens; cognitus synthesizes the full picture.
A web app to preserve and visualize personal relationships and shared memories. Store photos, audio notes, and text — see how everyone in your life connects through an interactive knowledge graph.
Real-time ambulance routing with live location sharing. OSRM for optimal routing, python-socketio for sub-second GPS updates, Leaflet.js for map rendering. Deployed on Render + Cloudflare Tunnel.
Real-time personal safety Android app — SOS triggers, live location sharing, and emergency alerts. Built in Kotlin with Jetpack Compose. When seconds matter, Aegis acts instantly.
Config-driven app generator — provide a JSON config, get back a fully scaffolded web app with UI, API, and database wired up automatically. Zero boilerplate, full-stack from a single file.
A collection of projects built while mastering full-stack web development with Python, FastAPI, and modern frontend tooling. The forge where everything started.
I'm a CS student at ANITS, Visakhapatnam, but I treat every project as a real system — not an assignment. I care about things that run in production, handle edge cases, and actually solve problems.
Current focus: AI systems with real infrastructure under them. Local LLMs, multi-agent orchestration, LLM routing, and privacy-first desktop AI. I want to build the layer that makes AI deployable without sacrificing trust.
One word on the GitHub bio sums it up: consistency. It's the only thing that compounds.
Looking for internships, open-source collaborations, and interesting problems. If you're building in AI, real-time systems, mobile, or full-stack — let's talk. Vizag → anywhere.