A young venture capitalist in Singapore has discovered that AI tools costing $120 per month can outperform a traditional intern, automating everything from meeting summaries to quarterly reports while freeing up time for higher-value work like investor meetings.

Marc Palet, 26, joined the venture capital industry straight out of college and has since invested in over 12 AI companies. As the only employee outside the United States at his firm, OMVC, he faced a familiar startup dilemma: hire support staff or find another way to manage the operational workload. His solution offers a revealing glimpse into how AI is reshaping knowledge work in 2026.

What Problem Was Palet Trying to Solve?

Palet's workflow involved constant information overload. He would read interesting articles on LinkedIn or X, only to forget them later. He recorded founder meetings but dreaded the manual transcription and note-taking that followed. Quarterly reports required sifting through months of scattered data. These tasks consumed hours each week, time he could have spent on deal sourcing and relationship building.

The turning point came when Palet discovered NotebookLM, Google's AI-powered research tool. He began creating repositories where he could paste interesting content from across the web. When writing blog posts or preparing for meetings, he could query these repositories and retrieve exactly what he needed. But this was just the beginning.

How Did He Build His AI Agent System?

A few months ago, Palet read an essay by Andrej Karpathy, a founding member of OpenAI, about building knowledge bases that automatically ingest and update information. Inspired, Palet built something similar for venture capital using Claude Code, an AI coding assistant from Anthropic. His system now pulls information from multiple sources and organizes it intelligently.

The system works like this: it ingests data from Gmail, Slack, call transcription services, newsletters, YouTube podcasts, and other sources where Palet encounters portfolio company updates. An AI agent parses this raw information and organizes it into structured pages for each portfolio company. Each page contains sections for fundraising, financials, and commercial updates. When a new Slack message arrives, the AI determines which company it relates to and updates the relevant page, rewriting content to maintain coherence rather than simply appending new text.

Ways to Automate Venture Capital Operations With AI

• Meeting Summarization: Record founder meetings using transcription services like Fireflies, then use AI to automatically generate structured summaries covering the company's problem, solution, market size, and other due diligence details without manual note-taking.
• Portfolio Knowledge Management: Use tools like NotebookLM to create repositories that collect and organize information from LinkedIn, newsletters, podcasts, and other sources, making it instantly searchable when writing reports or preparing for meetings.
• Quarterly Reporting: Deploy AI agents that automatically summarize portfolio activity across all companies during a reporting period, generating quarterly limited partner reports without manual compilation from multiple sources.
• Content Creation at Scale: Use AI agents to draft LinkedIn posts and deep-dive articles, increasing output from one article every two months to one every three weeks while maintaining quality and driving deal flow.
• Audit and Compliance Work: Automate the gathering and organization of information from portfolio companies for audit purposes, reducing the time spent on administrative tasks that don't directly generate investment returns.

What Results Has This Approach Delivered?

The impact has been measurable. Palet increased his LinkedIn content output from one deep-dive article every two months to one every three weeks. Each article brings in additional deal flow, and because he spends less time on operational work, he has more time for in-person meetings with founders. His quarterly limited partner reports now generate themselves by summarizing everything that happened across the portfolio during the reporting period.

"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. AI plus me is more powerful than me plus an intern," said Marc Palet.

Marc Palet, Venture Capitalist at OMVC

Palet's monthly AI subscription costs exceed $120, covering Claude, ChatGPT, n8n (a workflow automation platform), and various transcription APIs. He notes he would be willing to spend more if it increased his productivity further. This is a deliberate choice: rather than hiring an intern at a typical salary of $30,000 to $40,000 annually, he invests in AI tools that scale with his needs and don't require training or management.

Why Does This Matter Beyond Venture Capital?

Palet's experience reflects a broader shift in how knowledge workers approach productivity. He started using ChatGPT in 2023 but found the real breakthrough came with coding-focused AI tools like Claude Code, Cursor, and Replit. These tools allowed him to build custom solutions tailored to his specific workflow rather than adapting to off-the-shelf software. His approach suggests that the future of AI in professional services may not be about replacing workers entirely, but about augmenting individual contributors with intelligent automation that handles repetitive, high-volume tasks.

For other VCs, founders, and knowledge workers facing similar constraints, Palet's tech stack offers a template. The combination of transcription services, AI coding assistants, and knowledge management tools creates a system that learns from your work patterns and adapts to your needs. As AI capabilities improve and pricing potentially decreases, this model may become standard across professional services, consulting, and other knowledge-intensive industries.