Hi, I'mJash Kakadiya
Founder of DNJ Techx | CEO at YRAM BIOSOLUTIONS LLP | Building AI, SaaS & Automation Solutions with Python & Generative AI
About
Engineer building AI for healthcare and education
I'm Jash, an AI/ML engineer and founding partner at DNJ Techx. I architect backends and AI pipelines for real-world products — from clinical reasoning assistants to regulatory copilots and medical study tools. I focus on shipping secure, scalable systems where the AI layer actually moves the needle for users.
My toolbox spans Python, Django, PostgreSQL, GPT-based LLMs, LangChain, Next.js and AWS. I care about clean architecture, performance, and making complex things feel obvious.
Education
- B.E. in Computer EngineeringGovernment Engineering College, BhavnagarCGPA: 8.63
- HSCAshadeep Science BhavanGujarat Secondary and Higher Secondary Education Board
Skills
A full-stack AI toolkit
Drag the sphere or let it spin. Every node is a tool I use in production.
Languages
AI/ML
Backend
Cloud
Tools
Projects
Things I've shipped
AI products powering medical education, clinical reasoning and global regulatory workflows.
GastroAGI
AI Gastroenterology Learning & Clinical Support
Complete backend for an advanced AI platform for gastroenterology education and clinical support — real-time AI responses, personalized learning flows, and structured medical content for students, clinicians and researchers.
- Architected backend & DB with Django + PostgreSQL
- Streaming AI responses and chatbot pipeline for medical Q&A
- Integrated fine-tuned LLMs for medical accuracy
- Role-based access, subscriptions, performance APIs
- Deployed on AWS with CI/CD automation
LEXAMEDAI
AI-Powered Global Regulatory Assistance Platform
Backend infrastructure for an intelligent regulatory assistant supporting pharma and medical device companies through IND submissions, Phase I–IV trials and compliance with FDA, EMA, MHRA, CDSCO, PMDA, TGA and Medsafe.
- Backend architecture & DB model for structured regulatory content
- AI-driven Q&A and regulatory workflow automation
- Secure REST APIs + Next.js frontend with auth
- Reusable components for multi-region regulation comparison
- Cloud + CI/CD pipelines on AWS
MBBSAGI
AI Medical Learning Assistant for MBBS Students
Intelligent learning platform that simplifies complex medical concepts for MBBS students and exam aspirants. Real-time AI explanations, professor-style responses and personalized paths for NEET-PG and USMLE.
- Backend services & relational schema with Django + PostgreSQL
- Streaming AI responses and structured explanation workflows
- NLP-based topic classification and personalized learning logic
- Secure REST APIs and frontend integrations
- Optimized performance and scalability for student usage
YouTube Summary Extension
Chrome extension for AI video summaries
A Chrome extension that generates concise summaries for YouTube videos. Lightweight Flask/Django backend, GPT-3.5 powered summaries, packaged for the Chrome Web Store.
- Chrome extension UX integrated with YouTube
- Python backend handling transcript + summarization
- GPT-3.5 powered summarization pipeline
Product Image Matching
Visual similarity search using Inception V3
Built a product image matching system using the Inception V3 deep learning model, trained on a custom dataset of 3.6 million images sourced from ChromaDB for high-accuracy visual similarity search.
- Implemented Inception V3 CNN for image classification and feature extraction
- Trained on 3.6M product images from ChromaDB
- Built efficient similarity matching pipeline for real-time product lookup
- Optimized model inference for large-scale image retrieval
Experience
Where I've built things
Building AI-driven products and software solutions, leading product direction and engineering across multiple platforms.
Driving operations and strategic growth, with a focus on healthcare and biotech innovation.
Developed AI features and integrations as part of the core engineering team in Surat.
Built production AI/ML pipelines and Django backends serving real-world clients out of Ahmedabad.
Contact
Let's build something unreal.
I work on AI products, healthcare platforms, and ambitious ideas that need to actually ship. If that sounds like you, drop a line.