This as-told-to essay is based on a conversation with Marc Palet, 26, a Singapore-based venture capitalist at OMVC. The following has been edited for length and clarity.
I was first exposed to venture capital while running an e-commerce fashion business in college back home in Barcelona. During a study abroad program in Hong Kong, I did a ton of networking and came across a VC firm investing in fintech, blockchain, and AI — all the things I liked. I ended up staying with them after graduation and spent three years in Hong Kong before moving to Singapore to lead the firm’s Asia Pacific investments.
I’m 26 now and have invested in over 12 AI companies since I started my VC career. AI isn’t just a part of my portfolio; it’s become a crucial part of my workflow and how I manage things as the firm’s only employee outside the US.
Even though I’ve vibe-coded many of my workflows, AI hasn’t replaced any of my software subscriptions. It’s actually added more, since I pay more than $120 a month across Claude, ChatGPT, n8n, and other AI tools like transcription APIs.
But I’d be happy to spend more if it makes me more productive.
Here’s what my tech stack looks like and how it has helped shave countless hours of operational work.
Summarizing founder meetings
I started using ChatGPT in 2023. But the real shift came when I began using coding tools like Claude Code, Cursor, and Replit.
One of the first things I built for work automatically summarizes my meetings with founders.
We record our meetings using Fireflies. Instead of manually listening back and writing investment notes after every meeting, I receive a structured summary covering the company’s problem, solution, market, and other information we capture during due diligence.
My first knowledge base
One recurring problem was that if I weren’t at my laptop, I’d read something interesting on LinkedIn or X and then forget all about it.
So creating repositories on NotebookLM was a turning point for me. Every time I read something online that I thought I wanted to remember, I’d copy and paste it into that repository.
Whenever I wanted to write a blog post, I could query it and retrieve all the information I needed.
Building a company brain
The stage I’m in now started a couple of months ago, after I read something OpenAI founding member Andrej Karpathy wrote about building a knowledge base that automatically ingests information and keeps updating itself. I thought that was really interesting, so I built something similar for venture capital using Claude Code.
It pulls information from Gmail, Slack, my call transcription service, newsletters, YouTube podcasts, and anywhere else I get knowledge or updates on our portfolio companies. An AI agent parses the raw information, then organizes it into the right place.
Each portfolio company has pages for things like fundraising, financials, and commercial updates. If there’s a new Slack message, the AI figures out which company it’s about and updates the relevant page. And it doesn’t just copy and paste new information — it rewrites the page so everything stays coherent.
More powerful than an intern
One workflow automatically creates our quarterly limited partner reports by summarizing everything that’s happened across our portfolio during the reporting period. Another helps with audit work by gathering information from all our portfolio companies.
Those tedious tasks that used to take up so much of my time. For a VC fund as small as ours, time is always a constraint, so I’m happy to spend on AI that can save me time.
My agents for LinkedIn posts, an important part of my role, have helped me increase my output from one deep-dive article every two months to one every three weeks. Every article brings in more deal flow, and because I’m spending less time on operational work, I have more time for in-person meetings.
I realized that AI is much more efficient than training an intern for a month, only for them to stay for three months. AI also does things the way I want them done. I realized that AI plus me is more powerful than me plus an intern.