---

> whoami

name: Dhevenddra K G
location: Chennai, India
role: Development Engineer 2 @ Comcast — AI-native supplier onboarding platform
education: B.Tech CSE (AI & ML) — SRM University — CGPA 9.26
languages_spoken: [Tamil, English, Hindi]

currently_building:
  - forensic-deepdive — code knowledge graphs for AI agents (PyPI)
  - agentic-e2e — AI end-to-end testing for pull requests (npm)
  - Tanglish voice agents and from-scratch speech pipelines

interests:
  - LLM agents, RAG, MCP tooling
  - voice and speech systems for Indian languages
  - making non-deterministic systems auditable
  - creative filmmaking

At work: LLM workflows inside a compliance process, where every AI-assisted decision has to be explainable after the fact.
Outside it: tooling that makes AI agents useful on real codebases. 237+ production commits, 9+ clinical scales running in hospitals, two packages published. ↓

---

⚡ Tech Stack

AI / ML

LANGUAGES

BACKEND & DATA

FRONTEND & MOBILE

TOOLS & INFRA

---

🔨 What I've Built

🧬 forensic-deepdive Code knowledge graphs for AI agents PUBLISHED Python Static analysis Graph modeling Apache-2.0 Multi-language static-analysis engine that turns a codebase into a queryable knowledge graph — with a per-edge confidence taxonomy (EXTRACTED / INFERRED / AMBIGUOUS), git-history archaeology modelled as graph data, and automated architecture docs across six output surfaces including Obsidian vaults. Built so coding agents can reason about unfamiliar code instead of guessing at it. v0.8, shipped across an ADR-driven roadmap with 90+ logged decisions.	🧪 agentic-e2e Plain-English E2E tests, run by an AI agent PUBLISHED TypeScript LLM agents Playwright/CDP GitHub Actions Zod Write the test in English; an agent runs it against a live preview of every pull request. It reads the page's accessibility tree, caches what works so steady-state runs cost zero LLM calls, and re-resolves only the step that drifted when the UI changes. Fork-safe two-workflow model lets a fork's PR post a comment and status check without ever touching repo secrets. Ships with a flake-vs-regression classifier and watchable failure replays.
🗣️ VoiceDesk Tanglish voice service-desk agent Python Agentic workflows BM25 RAG MCP ASR/TTS Provider-agnostic voice agent with typed ports and a zero-key local mode. BM25 retrieval, working plus long-term memory, guardrails, MCP-shaped tools, and a draft → confirm → execute flow that stops the agent making state changes nobody approved.	🎙️ VoiceGate Streaming VAD & ASR front end, from scratch Python NumPy SciPy scikit-learn Whisper/Saaras Voice-activity detection built from the DSP up — hand-written features, adaptive gating, smoothing, and a learned logistic-regression path — plus a mic/file-to-ASR pipeline with Whisper and Saaras adapters. 31 passing tests.
📈 MARKGAP Autonomous options trading agent HACKATHON Python Alpaca API Options Deadline-aware options agent that optimises realised cash rather than paper P&L. A deterministic engine makes every trade decision; the LLM is confined to risk-scaling and journaling. Built for the lablab.ai × Alpaca hackathon.	⚖️ CLAUSE Auditable AI claims adjudication HACKATHON FastAPI Gemma vLLM/ROCm AMD MI300X APPROVE / FLAG / DENY decisions grounded in verbatim policy clauses, with every citation programmatically verified against the source before display — failed citations are visibly rejected, never silently dropped. One structured-output call, switchable between Fireworks AI and a self-hosted MI300X via one env var.
🐾 agentrocky-win Pixel-art companion that runs Claude Code on Windows Electron TypeScript React Vite Windows port of @itmesneha's macOS agentrocky. Rocky walks along your taskbar, opens a retro terminal chat when clicked, and runs Claude Code as a persistent subprocess. The Windows-specific work was in the corners: claude.cmd resolution, transparent overlay quirks, stream-JSON CRLF tolerance, DPI-aware sprite rendering.	🏥 ClinicOS Multi-tenant clinic management SaaS Next.js TypeScript Supabase Tailwind End-to-end practice management — patient records, appointments, billing, multi-clinic tenancy. Built with Claude Code as the engineering pair. The story site walks through the build approach and the design decisions.
♟️ ChessText Real-time multiplayer chess — with text-based moves React TypeScript Supabase WebSockets Move pieces by typing e2e4. Real-time sync, move validation, time-limited hints, persistent profiles. Built because I wanted chess without dragging pieces.	📡 Iris Nearby P2P messaging with zero internet Flutter Dart Bluetooth WiFi Direct Android app for peer-to-peer messaging over Bluetooth and WiFi Direct. E2E encrypted, auto device discovery (100m range), voice notes, images, offline message queuing.

🗂️ More projects

Project	Stack	What it does	Link
NCRP Reimagined	Next.js, Vercel	Cybercrime reporting portal redesigned as a guided, accessible flow	Live ↗
InsightOS	Python, Next.js, HuggingFace	Summarization, emotion detection, SEO scoring, toxicity analysis, rewrite modes	Repo ↗
RAG Chatbot	FastAPI, ChromaDB, Ollama	Document Q&A with semantic chunking, vector search, local inference	Repo ↗
Voice Assistant	Python, Speech APIs	Real-time speech-to-speech, latency-optimised, local models	Repo ↗
GreenLedger	Electron, Node.js, AI/NLP	Loan agreements → live ESG dashboards, all inference local	Devpost ↗
Web-Vex	Svelte, TS, Web Workers	Terminal-style spreadsheet viewer for CSV/XLSX	Live ↗
ListPad	React, JS	Minimal, login-free real-time collaborative lists	Live ↗
BikeBook	Web	Motorcycle service records & maintenance scheduler	Repo ↗

---

📊 GitHub Stats

---

---

🤝 Let's Connect

Happy to talk about LLM agents, retrieval, voice systems for Indian languages, distributed backends — or filmmaking.

If you're hiring → I'm looking for AI and backend engineering roles. LinkedIn ← let's talk.