Show HN: Medullar, the AI-Powered Data Discovery and Insight Platform
medullar.comHi HN,
I want to share a project I’ve been working on for the past year: Medullar. It’s a platform for teams to search, organize, and extract actionable insights from documents, files, and conversations across many different tools.
*Background:* I started this after years of frustration managing files and discussions scattered across cloud drives, email threads, chat logs, and more. Searching for the right bit of knowledge in Slack, Google Drive, or Dropbox was a daily pain, especially as projects scaled and teams grew. Existing “unified search” products either ignored security, required moving data around or lacked actual collaborative spaces for analysis. This made knowledge discovery slow and sometimes incomplete.
*What Medullar does:* * Federated, AI-powered search (NLP-based, not just keywords) across 60+ connectors (e.g., Google Drive, Slack, Dropbox, Outlook, Salesforce). * Everything is surfaced “in place”: we don’t move your data unless you explicitly choose to import it. * Spaces: collaborative environments (think project “rooms”) where teams pull in relevant docs, emails, and chats, annotate, extract, and discuss insights, and build up a living knowledge base, all with granular access controls and encryption layered in. * The AI helps interrogate, summarize, and connect ideas within and across files.
*What’s different:* Unlike other tools, you don’t lose privacy; there’s end-to-end encryption and zero data movement by default. You aren’t just collecting files but organizing and sharing insights among teams, which helps keep things actionable. You can search across your entire tool landscape in natural language, not just filter by keywords.
*Trying it out:* Anyone can go to https://www.medullar.com and start a 30-day free trial (no credit card required). If you email me with honest feedback or a bug report, I’ll extend your trial by another month. This is a genuine ask for feedback.
I’m happy to share details or answer questions if you want to know more about the architecture, limitations, or our federated query process. Feedback, especially rough edges, developer pain points, or skepticism about our privacy claims, is appreciated.
Thanks for reading! I’m looking forward to your thoughts.
What do you mean by "you don’t lose privacy; there’s end-to-end encryption and zero data movement by default"?
How does your NLP/ML work if it's encrypted?