Dossium

Great article. This is exactly what we’ve built here with @dossium Hearing the same thing across so many verticals - everyone needs this layer built, for themselves and for their agents.

With a new release of @dossium comes an updated, sleek website - thanks to @GeminiApp for being a tasteful design partner.

Had the perfect @dossium experience today. Met with a VC this morning, and as I was pitching, I explained our core value in an elegant way I don’t think I’d said before. But after the meeting I couldn’t remember the exact words. Thankfully I had my meeting recorder connected to @dossium and it already ingested the transcript after the meeting. So I opened the transcript page, went to the “content chat” panel and asked it to extract that part of my pitch in a clean way. Seconds later, 💥 perfect concise pitch I can drop into emails or use in our investment memo.

Things are getting wild. Asked our @dossium agent to analyze its content ingestion timeline of the last month. It called our 'query_help' tool to codegen a call to the internal SDK, and executed it on the embedded @daytonaio sandbox. Had it return an image artifact (which gets auto-ingested for later search). This kind of analysis is amazing - and just within a min or two - using your preferred LLM, and fully customizable via prompting the agent.

Any LLM client like @claudeai or @ChatGPTapp can show you your chats and let you search them. Only @dossium auto-summarizes, extracts entities and facts, and creates a context graph to see the relationships between your chats.

Hello world.