kevinbadi

kevinbadi / kevs-obsidian-ingestion-engine

Public

Generic AI-powered video content ingestion engine. Dropbox → Whisper → Claude → Obsidian knowledge graph + Postgres

10
5
100% credibility
Found Apr 14, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

This repository provides a configurable engine that downloads videos from Dropbox, transcribes and analyzes them with AI, classifies content using a custom taxonomy, generates detailed markdown notes for Obsidian, and indexes metadata in a Postgres database.

How It Works

1
🔍 Find the Video Organizer

You discover a helpful tool that transforms your pile of unprocessed videos into searchable, connected notes in your note-taking app.

2
📦 Set It Up on Your Computer

You bring the tool to your computer and make sure everything is installed and ready to start helping you.

3
📝 Tell It About Your World

You describe your project by listing video types, topics, people, locations, and moods that matter to you.

4
🔗 Connect Your Favorites

You link your video cloud folder, thinking helper, note collection spot, and search storage so they all work together seamlessly.

5
📁 Choose Your Video Folder

You pick the folder full of videos you want to organize, and give the go-ahead.

6
⚡ Let the Magic Happen

The tool grabs each video, listens to the words, spots key moments, understands the content, and builds smart notes – you watch it all unfold.

🎉 Dive Into Your Knowledge Garden

Your note app now blooms with rich pages featuring thumbnails, quotes, summaries, smart links between people and topics, and everything searchable forever.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 10 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is kevs-obsidian-ingestion-engine?

This JavaScript-based, AI-powered ingestion engine automates turning raw video files from Dropbox into searchable knowledge in Obsidian and Postgres. Point it at a folder via CLI, define a JSON taxonomy for your content types, topics, people, and tones, and it transcribes with Whisper (local or API), extracts key frames, analyzes visuals with Claude, classifies videos, then generates rich Markdown notes with thumbnails, quotes, clips, wiki-links, and frontmatter—while indexing metadata for queries. It solves the problem of unsearchable video piles by building an auto-linked graph in minutes per file.

Why is it gaining traction?

Unlike generic GitHub tools or basic transcription scripts, it's fully config-driven with zero code changes needed for custom taxonomies, handles parallel batches, skips duplicates, and cleans up temp files to save disk space. Developers dig the end-to-end pipeline—real-world tested on 900+ videos—delivering Obsidian-ready notes with embeds and a queryable DB, plus easy source plugins for Dropbox (Google Drive next). The Claude integration for vision and classification feels polished without vendor lock-in.

Who should use this?

Content ops leads at video-heavy orgs like clinics or podcasts managing Dropbox archives. Obsidian enthusiasts building personal knowledge graphs from interviews or b-roll. Indie devs or researchers automating ingestion for AI-powered content analysis without building from scratch.

Verdict

Grab it if you've got a video backlog and Obsidian setup—thorough README and MIT license make setup straightforward, despite 10 stars and 1.0% credibility score signaling early-stage maturity. Fork and extend for production; skip if you need battle-tested scale or face recognition today.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.