Back SAIM.
Case Study — 03

BIMO

A full-stack, end-to-end streaming AI chat workspace. Complete with single-click OAuth, multiple AI models with vision, real-time token streaming, and an analytics dashboard.

Year 2026
Role Full-Stack Developer
Stack Vanilla JSFlaskSupabaseRender
bimo.qzz.io
BIMO desktop preview
BIMO mobile preview

Overview

BIMO isn't just a wrapper; it's a private, comprehensive conversation lab. I designed this cloud-native application to give absolute control over the AI chatting experience seamlessly. We have three modes: Fast, Thinking & Coding, and Deep Reasoning for complex problems.

Built with a "no framework" pure HTML/JS frontend making it incredibly fast. Everything connects to a scalable Flask Python gateway deployed on Render via Server-Sent Events (SSE) that ensures token-by-token streaming right into the UI. No spinners, no lag. Supabase handles Row-Level security so every chat session and attachment upload is safely scoped to its user.

With features like syntax-highlighted code execution markdown, structured feedback data generation for the built-in analytics dashboard, and OAuth Google Sign-in backing it, BIMO represents a robust, decoupled, and production-ready full-stack application.

Preview Github