Sujith's AI Twin
Welcome to my neural interface. I am the AI Digital Twin of Sujith Senthilraj. Ask me anything about my journey, skills, projects at Prayag.ai, or coding experience.
Does Sujith fit for your role?
Career Journey
Sujith's StorySynthesize a fluid, storytelling narrative of Sujith's journey, drawing connections between his academic background, self-learning, and professional RAG work at Prayag.ai.
Skills & Expertise
Categories
AI / ML & Agents
Core specialization in Large Language Models, RAG architectures, and Multi-Agent Orchestration.
Project Showcase
PMT – MCP Conversational Agent
Project management platforms with 40-60 APIs are complex to navigate, requiring users to learn custom UIs and manually click through workflows.
Knowledge Base Chatbot
Flat vector embeddings lose structural and dependency relationships in complex codebases or internal documentation.
Property Document Analyzer
Analyzing hundreds of legacy property documents for legal links and ownership transfers is manually slow and error-prone.
EasyHire.ai – AI Hiring Assistant
Recruiters spend hours manually scanning hundreds of resumes, leading to screening bottlenecks.
Nanban Fund
Tracking monthly contributions, payments, and defaulters in small friend-circle chit funds is tedious when managed via Excel/paper.
Salary Prediction
Providing fair salary offers based on subjective interview scores and candidate profiles without statistical backing.
Career Timeline
Junior AI Engineer (formerly AI Engineer Intern)
Prayag.ai
- Architected a FastMCP server exposing 15+ project management APIs as MCP tools, enabling an OpenAI-powered agent to dynamically select and invoke endpoints based on user intent — eliminating hardcoded routing logic entirely.
- Designed a dual-source RAG pipeline combining ChromaDB vector embeddings (for documentation) and a Neo4j graph database (for codebase relationships), improving retrieval accuracy over flat-embedding approaches.
- Built a 3-agent orchestration pipeline (intent classifier → hybrid retriever → response synthesizer) to deliver grounded, context-aware answers from heterogeneous knowledge sources.
- Engineered an async document intelligence system using Redis-backed dual workers — one for AI embedding ingestion, one for Neo4j graph construction — processing 500+ legacy property documents with relationship extraction across configurable rule sets.
- Integrated speaker diarization pipelines using Pyannote.audio and AWS Transcribe to extract per-speaker analytics from audio recordings for enterprise use cases.
- Implemented OpenAI embeddings and vector storage workflows to power semantic search across large enterprise knowledge bases.
B.E in Computer Science and Design
Karpagam College of Engineering (Anna University)
- Score achieved: 8.05 CGPA
LeetCode & Competitive Programming
Key Achievement
- Solved 240+ problems on LeetCode (earned the Pandas badge) and 300+ problems across multiple competitive coding platforms.
Java (Netstack)
Issued by Netstack
- Successfully certified in domain capabilities.
JavaScript (IIT Bombay)
Issued by IIT Bombay
- Successfully certified in domain capabilities.
Team Lead — Sustainability Hackathon
Hackathon Event
- Developed a web-based environmental sustainability solution, managed timelines, coordinated task execution, and presented the final prototype to the judges.
2nd Place — UI Design Competition
UI/UX & Frontend Challenge
- Earned second place in a competitive UI design challenge at Kongu Engineering College, outperforming 50+ participants with creative and user-centered designs.
Python (GUVI)
Issued by GUVI
- Successfully certified in domain capabilities.
Introduction to AI (Novitech)
Issued by Novitech
- Successfully certified in domain capabilities.
HSC in Science (Computer Science & Mathematics)
KSV Matric Higher Secondary School
- Score achieved: 84%
SSLC in General Schooling
Mount Giris Matric Higher Secondary School
- Score achieved: 94%