Skip to content
View kunwardhruv's full-sized avatar
💭
busy
💭
busy
4 followers · 14 following

Highlights

  • Pro

Block or report kunwardhruv

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kunwardhruv/README.md

Typing SVG

Profile Views   Open to Work   Academic   Location

LinkedIn Email GitHub Portfolio Resume

🧑‍💻 About Me

"I don't just use AI — I build things with it."

I'm a B.Tech AI & Data Science student at GGSIPU, Delhi, deep in the Agentic AI space — building autonomous systems that reason, retrieve, and act. Across my projects I've gone from single-agent RAG chatbots to full multi-node LangGraph pipelines that take a raw dataset or document on one end and return a fully reasoned, explained result on the other — with almost no manual glue code in between.

I care a lot about why a system makes the decisions it does, which is why explainability (SHAP, page-cited RAG, structured LLM reasoning) shows up in nearly everything I build.

🎓 Background

  • B.Tech in AI & Data Science @ GGSIPU, Delhi (2022 – 2026) · CGPA: 8.4
  • 2 published research papers — Sports Analytics (Big Data + Deep Learning) & AI Resume Screening (NLP + Transformers)
  • Hands-on industry exposure across computer vision, IoT, medical imaging, and LLM-driven R&D

📌 Open To

  • ⚡ Immediate joiner — AI/ML Engineer & Agentic AI / LLM Engineer roles
  • 🧭 Open to relocating anywhere in India
  • 💬 Always up for a chat about LangGraph, RAG, or LLM orchestration

💼 Experience

🏢 HRS Navigation — R&D AI/ML Intern 📅 Jan 2026 – Apr 2026 · Bengaluru
  • 🧠 Engineered context-aware prompt workflows using LLMs to automate structured information extraction from technical documents
  • ⚙️ Built a RANSAC-based circle & blob detection pipeline on medical imaging data — improved robustness by 40% over baseline
  • 📄 Authored internal research report covering pipeline design, evaluation metrics, and deployment recommendations

🏢 QVOLV Technologies — AI/ML Intern 📅 May 2025 – Aug 2025 · Noida
  • 🎯 Built a YOLO-based object detection pipeline achieving 95%+ accuracy using PyTorch and OpenCV
  • 🗣️ Developed end-to-end speech-to-translation system using Flask REST APIs + HuggingFace MarianMT
  • 🔧 Implemented IoT predictive maintenance system using Arduino sensor data and time-series forecasting

🛠️ Tech Stack

Languages

Python JavaScript TypeScript SQL

AI / ML / DL

scikit-learn XGBoost PyTorch OpenCV Pandas

LLMs & Agentic AI

LangChain LangGraph Groq HuggingFace FAISS

Backend & Frontend

FastAPI Flask Next.js React Streamlit

Tools & Deployment

Git Vercel Render Supabase

🧠 AI / ML Expertise

Domain Level Details
Agentic AI (LangGraph) ⭐⭐⭐⭐⭐ Multi-node StateGraph pipelines, conditional routing, shared TypedDict state, multi-agent supervisor systems
RAG Systems ⭐⭐⭐⭐⭐ FAISS + BM25 hybrid search, Reciprocal Rank Fusion, hallucination guardrails, page-cited retrieval
LLM Integration ⭐⭐⭐⭐⭐ Groq Llama 3.3 70B, prompt engineering, multi-provider fallback chains, structured JSON output
Classical ML ⭐⭐⭐⭐ scikit-learn, XGBoost, automated model selection, GridSearchCV, SHAP explainability
Computer Vision ⭐⭐⭐⭐ YOLO real-time detection, OpenCV pipelines, RANSAC-based medical imaging
Full-Stack AI ⭐⭐⭐⭐ FastAPI backends, React/Next.js dashboards, SQLite/Supabase, WebSocket real-time

📊 GitHub Analytics

Contribution Graph

🎯 Current Focus

currently:
  building:
    - Real-Time Voice Translation pipeline (Groq Whisper → ElevenLabs TTS)
    - Autonomous Data Scientist Agent (7-node LangGraph + SHAP)
  learning:
    - Advanced LangGraph patterns (sub-graphs, human-in-the-loop)
    - Multi-agent orchestration at scale
    - Production RAG evaluation & tracing
  open_to:
    - AI/ML Engineer roles
    - Agentic AI / LLM Engineer roles
    - Immediate joiner · Open to relocation anywhere in India

🌐 Connect

LinkedIn GitHub Portfolio Email

"I don't just use AI — I build things with it."

Pinned Loading

  1. legal-contract-analyzer legal-contract-analyzer Public

    AI-powered legal contract analyzer using RAG — automatically detects risky clauses, explains them in plain English, and answers questions about your contract. Built with LangChain, Groq (Llama 3.3 …

    Python

  2. Autonomous-Data-Scientist Autonomous-Data-Scientist Public

    Autonomous Data Scientist Agent — Upload any CSV, agent automatically runs EDA, selects best ML model, tunes it, and explains results via SHAP + Groq LLM.

    Python

  3. Edurag-chatbot Edurag-chatbot Public

    RAG-based educational chatbot for Class 1-12 students. Upload any NCERT textbook PDF and ask questions — answers come strictly from the book with page citations.

    TypeScript

  4. medical-ai-assistant medical-ai-assistant Public

    AI-powered post-discharge care assistant — LangGraph multi-agent system with RAG, patient identification, triage assessment & real web search. Built with Groq + FAISS + FastAPI.

    Python

  5. Realtime-Voice-Translator Realtime-Voice-Translator Public

    Real-time AI voice translator — speak in one language, hear it translated live in another, with VAD-based chunking and streaming TTS

    Python 2

  6. Supervisor-Multi-Agent-Content-Team Supervisor-Multi-Agent-Content-Team Public

    Supervisor-orchestrated multi-agent content pipeline (LangGraph) with 4-way LLM fallback (Gemini/Groq/OpenRouter/Cerebras), bounded revision loop, and a Streamlit UI.

    Python