Education
9.04CGPA
B.Tech, Computer Science (AI and Data Science)
MIT World Peace University, Pune.
Web development certification
Training in modern web technologies.

AI Engineering. Models, pipelines and evaluation, taken from a notebook to a service that holds up in production.
©2026
9.04CGPA
B.Tech, Computer Science (AI and Data Science)
MIT World Peace University, Pune.
Web development certification
Training in modern web technologies.

Not just
code
02roles so far
Freelance developer
Independent
Web applications for clients, with API integrations and custom features: turning requirements into working software, designed and built end to end.
AI intern
Built AI solutions including chatbots, hospital assistants, document verification and automation tools, using Python, FastAPI, LangChain, OCR, YOLO and Docker.
I'm Tanish Bhavsar, an AI engineer and full-stack developer based in Pune. I build AI systems and the interfaces around them, from the model to the last pixel.
It started with curiosity about how things work. I began with plain HTML and CSS, moved on to JavaScript, React and the backend, and kept getting pulled toward the place where design and function meet.
My interest in AI grew alongside the web work, once I saw how much it could change the way people use technology. Now I do both: the model and the product around it. Most of what I know came from building real things, alongside my degree.
I'm Tanish Bhavsar, an AI engineer and full-stack developer based in Pune, India. I build AI systems and the interfaces around them, from the model to the last pixel. I'm studying for a B.Tech in Computer Science (AI and Data Science) at MIT World Peace University, Pune.
AI Engineering: Models, pipelines and evaluation, taken from a notebook to a service that holds up in production. Agentic Systems: Tool-using agents with guardrails, memory and clear failure modes, built on Gemini and Claude. Full-Stack Web: Next.js, TypeScript and Postgres. Fast pages, typed APIs and deploys that don't need babysitting. UI/UX: Interfaces designed in Figma and shipped in code, with the details intact. Hardware & IoT: Raspberry Pi and Arduino builds that put voice, sensing and AI into the physical world.
Claude, Figma, Framer, GitHub, Docker, Webflow, JavaScript, Photoshop, Python, C++, React, Next.js, FastAPI, Node.js, TensorFlow, OpenCV, Gemini, LangChain, Firebase, MongoDB, Blender, Raspberry Pi, ElevenLabs, TypeScript, PyTorch, PostgreSQL and Arduino. I pick the tool that fits the problem, and I'd rather go deep on a few than skim a hundred.
Rapid Crisis Response was selected as a Top 100 project in the Google Solution Challenge 2026. AI emergency coordination for hospitality venues, from panic button to staff brief. Live work includes CredZen, Ignisia '26, Ajinkya Jadhav Photography and FIFA World Cup 2026 Dashboard. There are 10 case studies on the work page, each with the problem, the approach and the stack.
Freelance developer, Nov 2024 – present. Web applications for clients, with API integrations and custom features: turning requirements into working software, designed and built end to end. AI intern at Stark Digital Media Services, Jun – Aug 2025. Built AI solutions including chatbots, hospital assistants, document verification and automation tools, using Python, FastAPI, LangChain, OCR, YOLO and Docker.
Yes. Open to freelance and internships. Email tanishbhavsar2007@gmail.com or use the contact form; I usually reply within a day.