Problem
Document review can be slow and repetitive, especially when users need to extract structured insights from files while keeping their data separated by account.
AI / Document Intelligence Prototype
Dokla
AI-powered document analysis app with user authentication and private analysis history.
Role
Full-Stack Developer
Year
2026
Stack
Next.js + Supabase
Solution
Dokla combines Gemini-based document analysis with Supabase authentication and per-user analysis history, storing structured analysis output instead of original documents.
Overview
What I Built
Dokla is a document analysis application built with Next.js, Supabase, and Google Gemini. It allows authenticated users to analyze documents while storing only structured JSON analysis results and metadata, with Row-Level Security protecting each user's data.
Technical Highlights
Engineering Notes
Next.js app router implementation
Supabase auth and protected user data
Gemini integration for analysis generation
Privacy-aware storage model for analysis metadata