Skip to content

🔍 Features

🔐 Authentication

DocGen.AI provides a secure and seamless authentication system with support for:

OAuth login via Google, Facebook, or Github

Email-based registration

Email verification for user to be activated

Two-Factor Authentication (2FA) for enhanced security (optional)

Email-based password reset

🧠 How It Works

  • DocGen.AI connects to your codebase or documentation context.
  • Users can chat with an LLM to generate unit tests, inline documentation, or ask questions about code behavior.
  • You can either connect a GitHub repo or chat with a blank model.
  • The system streams results in real time and provides copyable output.
  • If no codebase is loaded, DocGen.AI defaults to chat-only mode for exploration and experimentation.
Interactive AI Workspace Demo

👤 Guest User Functionality

DocGen.AI offers guest sessions for quick testing or onboarding:

  • Guests can explore model features without signing up
  • All guest data (code snippets, chats, preferences) is temporary
  • Guests cannot modify their profiles and have a limited time before their session expires (1 hour)
  • Session data is automatically wiped after logout or timeout

🖥️ Main Interface

Once logged in, users are welcomed into a responsive, modern workspace built for seamless interaction and model experimentation.

  • Toggle between sleek dark and light themes for a personalized experience
  • Intuitive dashboards to browse installed models and codebases
  • Real-time charts and usage metrics for insight into model/codebase activity
  • Full chat management: rename, edit, and delete conversations effortlessly
Demo

✍️ AI-Generated Code + Docs

The code generation panel allows you to request:

  • Unit tests for specific functions or classes
  • Inline comments for undocumented logic
  • Refactored or simplified code
  • JSDoc or docstring templates

📂 Codebase + Chunking Interface

Users can connect external repositories or upload local projects for embedding. DocGen.AI then:

  • Parses files and chunks content intelligently (e.g. by function)
  • Embeds code for retrieval-augmented generation (RAG)
  • Displays loading state and import logs
  • Supports reprocessing individual files for updates

🔐 Privacy + Local-First Deployment

DocGen.AI is designed for privacy-first use cases:

  • All processing can be done locally or in a containerized environment
  • Supports Ollama and other LLM backends for offline LLM inference
  • Codebases never leave your machine in local mode
  • Ideal for internal company projects and enterprise security needs Together, these features ensure that trust is not assumed—it’s provable.

🧠 Models

DocGen.AI supports multiple generative models via a modular architecture. Models can be used for chat, test generation, documentation, and search — with flexible backend integration via Ollama, OpenAI, or other compatible APIs.

Key features include:

  • Users can view a searchable list of all available and installed models
  • Usage statistics and last-seen data help track model engagement
  • Tags indicate whether a model is installed locally or just available to pull
  • Status indicators show loading, pulling, or ready state in real time

🔒 Admin-Only Installation

To maintain a secure and controlled environment:

  • Only admin users can pull or install models to the local environment
  • A model pull dialog allows admins to select specific versions or tags
  • Users without admin privileges will see a disabled or hidden pull button
  • Install requests are tracked and reflected in the UI for transparency

This ensures that the system maintains model consistency and avoids unintentional overload of the host machine.