I build AI-powered products that solve real business problems — from zero to production.
As a Founding Engineer at two startups, I've architected GenAI systems, optimized data pipelines at scale, and shipped backend platforms handling millions of daily operations.
- 🔭 Currently building at Cotelligent, A TechDemocracy Company — GenAI compliance systems, MLOps, and enterprise IAM analytics
- 🚀 Previously at Done Deal — AI-powered M&A platform with NLP, vector search, and secure data rooms
- 🎓 B.Tech, IIT Patna (2016–2020)
- 📫 Reach me at sujit.v.yadav@gmail.com
- 🌐 Portfolio: jitsu-13.github.io
@ Cotelligent (2025–Present)
- 🤖 GenAI enrichment layer converting complex IAM/permission data into compliance-ready narratives (~90% quality via RAG + LangChain + structured evaluation)
- ⚙️ End-to-end MLOps pipeline: human feedback → labeled datasets → fine-tuning → staged releases — reducing manual overrides by ~75%
- 📊 Explainable risk scoring engine with regulatory signals, SoD conflicts, drift monitoring, and anomaly detection
- 🚀 ETL optimization: 67% faster extraction (60→20 mins) via parallel batch processing + CDC-based incremental loads
- 🔧 Kafka event-driven processing + Redis caching (50–80% speedups) + Kubernetes deployments
@ Done Deal (2023–2025)
- 🎯 Recommendation engine using vector databases & NLP for intelligent deal-matching
- 🎙️ AI transcription & summarization — cut manual note-taking effort by 60%
- 🔐 Secure document management system (data room) with encryption & granular access controls
- 📦 Scalable microservices architecture (Java, Spring Boot, Hibernate) for M&A platform operations
Languages & Core
AI / ML / GenAI
Backend & Frameworks
Data & Infrastructure
Cloud & DevOps
- 🎓 Machine Learning Specialization — Stanford Online
- 🤖 Foundation of Generative AI — Udacity
- 📋 Foundations of Project Management — Google
