• 

  SunnySett.com

Inspiration

SunnySett was inspired by the challenges developers and startups face when adopting AI — from finding the right models to preparing datasets, fine-tuning, and deploying them. Most existing tools only handle one piece of the workflow, forcing users to jump between multiple platforms. We wanted to simplify this process and make AI adoption accessible to anyone with a browser — no DevOps complexity, no setup friction.

What it does

SunnySett is an all-in-one AI hub that lets users discover, test, and customize pre-trained machine learning models, manage datasets, fine-tune models with guided workflows, and deploy results instantly. By unifying every step of the AI development process into one browser-based workspace, SunnySett helps developers and teams go from idea to deployable intelligence quickly and effortlessly.

How we built it

We built SunnySett using a modern full-stack setup combining React and TypeScript on the frontend, Python (FastAPI) on the backend, and PostgreSQL with Pinecone for data and vector management. Our architecture is designed to be model-agnostic, integrating APIs from multiple open-source AI frameworks and providers. Everything is containerized with Docker and deployed on Vercel to ensure scalability, flexibility, and fast user access.

Challenges we ran into

Some of our biggest challenges included integrating multiple ML APIs into a single smooth interface, efficiently handling dataset uploads and versioning, and maintaining responsive performance during heavy model operations. We also had to carefully balance scalability, cost, and user experience to deliver a powerful yet intuitive platform.

Accomplishments that we're proud of

We’re proud to have built a live, production-ready web app that’s already being tested by over 100 developers. We successfully created a unified AI Playground that simplifies model discovery, dataset handling, and fine-tuning into one seamless experience. Our modular architecture supports various frameworks and model types, and we’ve built an intuitive interface that makes AI workflows accessible for both experts and newcomers.

What we learned

Throughout this project, we learned how important workflow unification is for developers working with AI. We gained insights into designing intuitive user experiences for technical audiences, managing performance across diverse APIs, and building scalable infrastructure. We also discovered how essential clear guidance and smart defaults are for helping users navigate complex AI tasks with confidence.

What's next for SunnySett

Next, we plan to expand SunnySett by adding community-contributed models and datasets, introducing a freemium-to-enterprise business model, and enabling custom fine-tuning pipelines and model hosting. We also aim to build partnerships within the AI ecosystem and continue growing our developer community as we move toward our first public launch.

Built With

• aws-apis-&-integrations:-openai-api
• fastapi-database:-postgresql
• figma
• gcp
• hugging-face-hub
• modal.com
• next.js-backend:-python
• ngrok
• pinecone-(vector-database)-infrastructure-&-deployment:-docker
• postman
• react
• replicate-api-other-tools-&-services:-github
• typescript
• vercel