Master Prince
Full Stack Developer & Applied AI Engineer based in Delhi. I build AI-powered web products, fast APIs, and automation systems for startups, agencies, and businesses that want software to save time and move faster.
Currently at M37 Labs building production AI systems for enterprise clients in India and Malaysia, multi-agent automation, NL to SQL search, AI recruitment, and brand intelligence tools. I also take on freelance projects: if you need an AI feature, a full stack web app, or an automation workflow built cleanly and shipped fast, I'm available.
6+
AI systems shipped
End-to-end from spec to production
1,000+
tasks automated daily
Multi-agent platform in live enterprise use
70%
faster hiring for clients
Resume screener with 95%+ accuracy
200K+
articles made searchable
Plain-English NL-to-SQL over a 3.6 GB DB
Work Experience
Full Stack Developer · Applied AI
M37 Labs · Delhi, India · Full-timeVoice-Controlled Multi-Agent Platform
An enterprise-grade autonomous multi-agent platform that automates workflows across email, documents, calendars, spreadsheets, presentations, and voice systems using a unified multi-LLM architecture. Its core innovation, the AgentBrain Pipeline, creates intelligent agents through autonomous research, context fusion, and dynamic identity synthesis instead of static prompting.
AI Developer Assistant
Claude API-powered tool routing to agents for code Q&A, refactoring, Git, and OCR screen analysis. Dual OCR pipeline. MongoDB chat history.
Media Monitoring NL Search
200K+ press articles searchable in plain English over a 3.6 GB MySQL DB. Redis caching: 5s → 2s response time.
Influencer Discovery Platform
AI-driven profile discovery via Selenium + SerpAPI. Async PDF & Excel reports via Celery and Redis. FastAPI + Next.js 16 dashboard.
Persona-Aware Chatbot
Detects user persona, retrieves via ChromaDB semantic search, generates personalised responses via OpenAI. 5-scraper knowledge base.
AI Recruitment Platform
100+ resumes screened in under 30s. 70% ML + 30% LLM hybrid. Real-time WebSocket updates, Zoom scheduling.
Brand Intelligence Platform
Automated daily digests, AI chat, audio transcription, and PowerPoint generation. Node.js + TypeScript + FastAPI. W&B Weave LLM tracing.
DB & Infrastructure
PostgreSQL tuning + Redis caching API times down 40%, DB load down 60%. Docker Compose, AWS EC2, Vercel, Render.
Full Stack Developer Intern
Luchkee Health · Delhi, India · InternshipHealthcare Platform
MERN stack app with JWT auth, RBAC, and encrypted patient data storage. Built and shipped features across the full stack independently as part of a small engineering team.
Projects
Mental Health AI System
I fine-tuned Qwen on my Mac using Reddit data to build a 14 class symptom classifier that doubles as a chatbot. Runs entirely locally nothing ever leaves your device. Handles everything through FastAPI.
Delhi PM2.5 AQI Forecasting
Wrote a neural network from scratch in NumPy to forecast Delhi's air quality. Trained on 6 years of real sensor data. Hit R² = 0.95 better than what Random Forest could do. Built it without any ML frameworks, just the basics.
Semantic Memory Engine
Wrote a memory system that takes long conversations and compresses them into a compact format. Fine-tuned Qwen2.5 to think directly from the compressed data no need to unpack it back to words. Saves about 10x the tokens. All local with Ollama.
Technical Skills
Languages & Frameworks
Data & Cloud
AI & Machine Learning
Education & Certifications
B.Tech – Computer Science & Engineering (AI)
Jamia Hamdard University, New Delhi
Graduated 2025
Four-year AI-specialised engineering programme at Jamia Hamdard, one of Delhi's central universities. The curriculum leaned heavily into applied machine learning, computer vision, and NLP alongside core software engineering which is why most of my personal projects are end-to-end systems rather than just notebooks.
Final Year Project: Intelligent Proctoring Detection Framework
Browser-based exam monitoring system using OpenCV, MediaPipe, and face-api.js to detect gaze deviation, head pose, multiple faces, and unauthorised objects in real time. Achieved 95%+ detection accuracy. Auto-generates a timestamped violation report with annotated frames after each session. Built with Flask as the backend and runs entirely in the browser no native app install needed.
Certifications
Data Analytics with Python
NPTEL · 2025
Pandas, NumPy, Matplotlib, Seaborn, and statistical analysis using real-world datasets. Issued by IIT faculty through SWAYAM.
Introduction to Database Systems
NPTEL · 2025
Relational models, ER diagrams, normalisation, SQL, transactions, indexing, and query optimisation. Issued by IIT faculty.