HithaBadikillaya
@HithaBadikillaya
Maintainers
January 15, 2026 | Documentation In Progress

D.A.S.H

A self-hosted, privacy-first documentation platform for generating meeting minutes, captions, and letters from audio or video content.

Next.js Express.js Node.js Python Whisper FFmpeg documentation privacy self-hosted local-first student-tools open-source
View on GitHub Documentation
D.A.S.H

D.A.S.H

D.A.S.H (Document And Social Hub) is a self-hosted, privacy-first documentation platform designed to generate structured meeting minutes, social media captions, and formal letters from audio or video content.
Built with a local-first philosophy, D.A.S.H ensures temporary processing, template-driven workflows, and complete control over sensitive data.

Upload recordings from meetings, podcasts, or spoken content and convert them into clean, usable documentation without relying on third-party cloud storage.

Key Features

  • Meeting Minutes (MoM) – Automatically generate clean, structured meeting notes
  • Caption Generation – Create captions for Instagram, X (Twitter), LinkedIn, YouTube descriptions, and WhatsApp promotions with multiple tone options
  • Letter Generation – Generate formal letters, invitations, sponsorship requests, and follow-ups using customizable templates
  • Template-Based Workflow – Use predefined or custom templates to maintain consistent yet flexible outputs
  • Privacy-First Design – Files are processed temporarily and deleted automatically after completion
  • Self-Hostable – Fully deployable on local or private infrastructure with no vendor lock-in
  • User-Friendly Workflow – Simple upload → process → edit → export flow with progress indicators for long files

Tech Stack

D.A.S.H is built using modern web and AI tooling:

  • Next.js – Frontend framework for a fast and responsive UI
  • Tailwind CSS – Utility-first styling for clean and consistent design
  • Node.js & Express – Backend API and orchestration logic
  • Python Worker – Handles heavy AI processing tasks
  • Whisper – Audio transcription engine
  • LLM APIs – Intelligent content generation
  • FFmpeg – Audio extraction, preprocessing, and trimming

Project Goals

  • Simplify documentation from spoken content
  • Enable privacy-first, local-first AI workflows
  • Reduce manual effort in meetings and content creation
  • Provide an open-source base for experimentation and customization

Open Source & Community

D.A.S.H is an open-source project developed under the Sahyadri Open Source Community (SOSC).
We welcome contributors, learners, and open-source enthusiasts to explore and improve the project.

Contributing

Contributions are welcome.

Back to all projects