AI/ML Engineer · Builder · SRM Chennai → Prodapt Solutions
Turning data and models into things that actually work.
- 🎓 B.Tech in CS (AI & ML) from SRM Institute of Science and Technology — CGPA 9.04
- 💼 Currently working as a Software (AI) Engineer at Prodapt Solutions, Chennai
- 🔭 Building AI-driven pipelines, agentic systems, and backend tooling
- 🌱 Deep into LangChain, LangGraph, MCP-based orchestration, and Graph DBs
- 🧩 I like problems at the intersection of AI and real-world engineering — deepfake detection, RAG systems, agentic workflows
- 📫 Reach me at amaanmajid23@gmail.com
Languages
Backend Frameworks
Agentic AI & LLM Stack
ML / DL
Data & Infra
CS Fundamentals
Python · FastAPI · MCP
An agentic AI system using MCP-based orchestration to fully automate meeting workflows — scheduling, reminders, email coordination, transcript generation, and MOM creation using Google Calendar and Gmail APIs.
LangChain · LangGraph · FastAPI · FAISS · GPT
Voice-enabled RAG chatbot for university admission queries. Context-aware responses with 40% reduction in manual query handling effort.
rPPG · CNNs · OpenCV
Extracted rPPG signals from video using chrominance methods and trained CNNs on 4000+ videos — achieved 97% accuracy, outperforming traditional deepfake detection models.
Python · TensorFlow · scikit-learn · REST APIs
ML-based health monitoring system using wearable sensor data for early disease detection and personalized recommendations. Built for the University of Binghamton. 92% prediction accuracy.
| Role | Company | Impact |
|---|---|---|
| Software (AI) Engineer | Prodapt Solutions | Built Neo4j code graph pipelines (91.2% accuracy), reduced manual preprocessing by 90–100% |
| Tech & Services Intern | UKG | Built One-Stream — near real-time data sync for 3,000+ global businesses (99.97% faster) |
| AI Research Fellow | HyperVerge | Deepfake detector with 97% accuracy using STFT/FFT/DWT pipelines |
| ML Engineer Intern | GetSetYo | OCR passport scanner with 70% speed improvement; itinerary generator reducing effort by 60% |
Open to interesting problems in AI engineering, agentic systems, and applied ML. Let's build something.