Orbofi platform preview: agent discovery and listings.
01 / Orbofi
AI agents
Many AI tools. One clear screen.
Frontend development and backend contributions · Founding engineering team
Built Orbofi's frontend from scratch and later moved it from React to Remix to load data on the server. Added an agent directory that loads more results as people scroll, and brought several AI tools into one chat screen.
The chatbot builder needed approximately 35% less time to become ready for interaction.
- React
- Remix
- Redux
- Node.js
- MongoDB
- OAuth 2.0
Read the project storyFull story for Orbofi
The product
Orbofi lets people create and interact with AI agents. I worked on the main application screens, browsing agents, and the screens for using different AI tools.
What I built
As part of the founding engineering team, I built the frontend from scratch, including reusable screens, navigation, and managing information across the application. I also worked on user accounts and third-party sign-in with Node.js, MongoDB, and OAuth 2.0.
The challenge
After the React frontend was built, the product needed to load data on the server. It also needed to support a growing agent directory and several chat modes, including chat, image generation, website generation, and blockchain interactions.
How I solved it
I moved the frontend to Remix so the application could load data on the server. I built a staggered grid that loads more agents as people scroll. I organized the different chat modes in one screen while keeping their controls clear.
I displayed AI replies as they arrived instead of waiting for the complete response. I also reduced unnecessary screen updates, split code into smaller parts, and loaded features when needed.
The result
The platform launched with roughly 500 or more users. The chatbot builder became ready to respond to input in approximately 35% less time. The updated application could load data on the server and support the platform's growing set of AI tools.