DHANRAJ SHARMA

AI Automation Specialist · LLM Integration · RAG Pipelines · Workflow Automation

---

👋 About Me

I build AI automation systems that ship — RAG chatbots, LLM integrations, n8n workflows, and full-stack AI products that replace manual work with intelligent pipelines.

I don't just train models. I take ideas → build working products → deploy them.

Most recently: Completed an 18-week research internship at ISRO's Space Applications Centre where I built MOCK DRILL — a full-stack automated compliance audit platform that compressed a 4-day manual security audit into 4 hours (90% effort reduction). Received a formal Letter of Recommendation from the Head of Cyber Security & Network Division.

What I build for clients and teams:
→ RAG chatbots connected to documents, databases, or websites
→ LLM integrations (GPT-4, Gemini, Claude API) into existing products  
→ AI workflow automation using n8n, LangChain, and FastAPI
→ Custom AI tools, Streamlit dashboards, and AI pipelines
→ Full-stack AI applications (React + FastAPI + Docker)

I use AI tools to build fast — working product in days, not weeks.

---

🛠️ Tech Stack

AI & Automation LangChain, LangGraph n8n Workflow Automation RAG Pipelines Prompt Engineering Vector DBs (ChromaDB, FAISS) LLM APIs (GPT-4, Gemini, Claude)

Backend & APIs Python, FastAPI Celery, Redis PostgreSQL, MySQL Docker, Nginx REST APIs, JWT Auth GitHub Actions

ML & Deep Learning TensorFlow, Keras, PyTorch Scikit-learn, XGBoost OpenCV, MediaPipe HuggingFace Transformers NLP, Computer Vision MLOps, DVC, MLflow

Frontend & Cloud React, TypeScript Streamlit, Tailwind CSS Google Cloud Platform Azure Cognitive Services Kubernetes Linux, Git

---

🚀 Featured Projects

🔒 MOCK DRILL — Automated Security Audit Platform @ ISRO

Production system built at Space Applications Centre, ISRO

Automated compliance audit platform for critical network infrastructure. Consolidated 69+ security checks across DNS, email, web, and SSH/FTP into a single scan engine.

• Impact: Reduced 4-day manual audit to 4 hours (90% effort reduction)
• Stack: FastAPI · Celery · Redis · PostgreSQL · React · TypeScript · Docker · Nginx
• Features: CVSS v3.1 scoring · Auto-generated remediation playbooks · RBAC (4 roles) · Real-time dashboard · One-click PDF reports
• Recognition: Letter of Recommendation from Head of Cyber Security & Network Division, ISRO SAC

---

🤖 ClientSense — AI Lead Qualification & Follow-up Engine

Production-ready business automation workflow built with n8n and Gemini AI

AI-powered lead management system that automatically qualifies inbound leads, routes them into hot/warm/cold pipelines, sends personalized email responses, tracks lead engagement, creates follow-up reminders, and generates daily CRM follow-up digests.

• Stack: n8n · Gemini AI · Gmail API · Google Sheets · Google Calendar · Tally Forms
• Features: AI lead scoring · Automated email workflows · CRM tracking · Reply detection · Calendar automation · Daily lead digest
• Architecture: Event-driven workflow automation using webhooks, AI classification, conditional routing, and scheduled follow-up processes
• 🔗 Live Demo)

---

⛽ Fuel Route Optimizer — Full Stack AI Road Trip Planner

Live deployed product — Vercel + Render

USA road trip planner that computes the most cost-effective fuel-stop strategy for any route using a custom greedy look-ahead algorithm and live fuel pricing data.

• Stack: Django REST API · React · TypeScript · Mapbox GL · Docker
• Built with: AI-assisted development pipeline (Claude + Antigravity)
• 🔗 Live Demo

---

🤖 Simple RAG Pipeline

Modular LLM + Vector Search System

Clean, modular retrieval-augmented generation pipeline. Document ingestion → chunking → embeddings → ChromaDB storage → Groq LLM → grounded responses.

• Stack: Python · ChromaDB · Groq LLM · LangChain · Embeddings
• 🔗 Repository

---

🚗 AutoSage Vehicle AI — Multimodal AI Assistant

Computer Vision + Conversational LLM — SmartBridge Internship

AI assistant combining computer vision with Gemini API to analyze vehicle images and return structured insights through a conversational chatbot interface.

• Stack: Gemini API · Python · Streamlit · Computer Vision · Prompt Engineering

---

🌍 mula-sakshi-ai — Google Solution Challenge 2026

Hackathon Project — Team of 2

AI-powered solution detecting contradictions between citizen-reported evidence and official government records. Submitted to Google Solution Challenge 2026.

• Stack: TypeScript · Gemini API · Full Stack

---

🏥 Pancreatic Cancer Detection

Applied ML — Healthcare Domain

LightGBM classifier on medical dataset for early pancreatic cancer detection. Demonstrates applied ML in high-stakes medical decision-making.

• Stack: Python · LightGBM · Scikit-learn · EDA Pipeline

---

🏆 Achievements

Achievement	Details
🚀 ISRO Research Internship	Letter of Recommendation — Head of Cyber Security Division, SAC
🏅 Google Solution Challenge 2026	Participant — AI solution submitted as 2-person team
📚 Academic Excellence	CGPA 9.41/10 — B.Tech Computer Engineering (AI/ML)
💻 25+ Public Repositories	AI/ML, GenAI, Automation, Cybersecurity, MLOps
🎓 Class XII	91.2% — CBSE PCM

---

📜 Certifications

• Google Cloud — MLOps, Prompt Design in Vertex AI, GenAI Apps with Gemini & Streamlit
• IBM — Deep Learning Essentials, Getting Started with AI
• Microsoft — Computer Vision with Azure Cognitive Services
• Red Hat — OpenShift AI, System Administration I & II, Linux Fundamentals
• KodeKloud — RAG Crash Course, Docker for Beginners
• OpenCV University — OpenCV Bootcamp

---

🤝 Open To

✅ Freelance AI automation projects
✅ LLM integration contracts  
✅ Remote internships and full-time roles at AI-first startups
✅ Short contracts and one-off builds
✅ Collaborations on AI products

📧 hinatashoyo101824@gmail.com

---

Building AI systems that actually ship.