The dirty secret of angel investing is that most of the information that drives decisions isn't in pitch decks. It's in conversations. In podcast episodes where a founder mentions their traction before they're public about it. In industry reports buried in your inbox. In the mental model you built from talking to 15 competitors over two years and never wrote down.
I'm Liam Carter. I advise startups and make early-stage angel investments — roughly 8-10 new investments per year, plus ongoing portfolio work. At any given time, I'm actively tracking somewhere between 150 and 200 companies. The information I need to do that well doesn't arrive in a format any spreadsheet can capture.
My knowledge management strategy used to be: take notes in Notion, try to maintain a company database, rely heavily on memory. The result was that I had good knowledge about companies I'd invested in recently and increasingly fuzzy knowledge about everything else. I was making decisions on incomplete information not because the information didn't exist, but because I had no system for retaining and surfacing it.
The Information Problem in Early-Stage Investing
Early-stage investing is fundamentally a knowledge accumulation problem. You're building a picture of a market, a technology, and a team from dozens of fragmented sources over months or years. The quality of your decision depends on how much of that picture you can actually recall and synthesize when it matters.
The information arrives in every format: founder calls (which I record with permission), conference talks, podcast interviews, Substack essays, Twitter/X threads, industry reports, introduction emails. The common thread: almost none of it is searchable in a meaningful way without a significant processing step.
sipsip's Transcriber handles the audio and video layer. I record all my founder calls, export the audio, and process them through Transcriber. A 45-minute call becomes a clean, searchable transcript in about 3 minutes. But the transcript alone isn't what made the difference.
What Mindverse Added to My Investment Process
Mindverse runs distillation on everything I process. For a founder call, the distilled output includes:
- Key claims the founder made about their traction, market, and differentiation
- Open questions I didn't fully resolve in the conversation
- Red flags or inconsistencies that surfaced during the call
- Connections to other companies or market dynamics in my knowledge base
That last item is the most valuable. When I added a call with a B2B SaaS founder last month, Mindverse surfaced connections to three earlier conversations with competitors I'd spoken with over the past year. I hadn't made those connections explicitly — they'd been sitting in separate transcripts. Mindverse found them.
The Daily Brief added another layer. I subscribe to about 25 podcasts and newsletters relevant to my investment thesis areas. Mindverse processes new episodes and articles overnight and synthesizes the key signals each morning. Instead of spending 90 minutes reviewing sources, I spend 15 minutes on the brief and act on the two or three things that actually matter.
The Knowledge Management Strategy That Emerged
After six months of using Mindverse as my primary knowledge management system, my process looks like this:
Before a founder call: I query Mindverse for everything I've captured about the company, the market, and the founder. Takes 2 minutes and surfaces context I'd have otherwise missed.
During the call: I focus on the conversation. I record it.
After the call: I upload the recording. Transcriber and Mindverse process it. Within 30 minutes I have distilled notes with the key claims, questions, and connections — better than anything I'd have written manually.
Weekly: I review Mindverse's surfaced connections. Things I captured two months ago that relate to something I looked at this week. This is where a lot of the pattern recognition actually happens.
Portfolio work: When a portfolio company asks about a market dynamic or a competitor, I query Mindverse. I've captured more than I remember, and the distilled layer surfaces it faster than I could reconstruct it.
What Changed in My Decisions
I don't have a clean A/B test for this, but I can point to specific instances where knowledge I'd captured months earlier — and would have forgotten under my old system — surfaced through Mindverse and changed how I evaluated a company.
The most concrete case: a founder in a crowded market claimed their pricing was differentiated. Mindverse surfaced a call from 14 months earlier with a company in the same space that had tried the same pricing approach and explained in detail why it hadn't worked. I had that conversation. I didn't remember the specifics. Mindverse did.
That's the knowledge management strategy I'd been trying to build manually for years. Turns out it needed AI to actually work.
Related: AI Knowledge Management Tools in 2026 Complete Guide: Knowledge Management: The Complete Guide for 2026
Start building your knowledge base at sipsip.ai — the free tier is a good place to test whether the pipeline works for your specific inputs.
As an angel investor, I consume enormous amounts of fragmented information. sipsip Mindverse became the knowledge management strategy I'd been trying to build manually for years.
