Actively investing · $5M–$50M checks · Seed onwards

From chatbots to colleagues.

I invest in AI co-workers and the infrastructure that supports them, as General Partner & Head of Tech Investing at B Capital. I've also sat in your seat: 4× founder, then COO & CRO at Weights & Biases as it scaled from $3M to $50M ARR.

Building AI co-workers? Let's talk Read the thesis
Yan-David Erlich
Yan-David “Yanda” Erlich GP & Head of Tech Investing · B Capital
$550M+
Deployed into AI across B Capital, Coatue, and angel investments
55×
DPI on my Weights & Biases seed angel check
$3M→$50M
ARR at Weights & Biases as COO/CRO. Hit/beat plan every quarter
4×
Venture-backed founder. 3 acquisitions (LinkedIn, CAI, Qualcomm)
Investment Thesis

The Shift From Agents to Co-Workers

  1. 01 · Tools “Help me” Assist a person with a task.
  2. 02 · Agents “Do this for me” Execute a task end to end.
  3. 03 · Co-workers “Own this with me” Own outcomes alongside you. Where I invest

The difference between an agent and a co-worker is fundamental: co-workers retain context across interactions, develop expertise in your specific business, and operate autonomously within defined boundaries. They don't just respond to prompts; they anticipate needs, learn from feedback, and improve continuously, much like a great new hire.

The agent→co-worker transition is happening now. I invest in the applications (engineering, sales, finance) and infrastructure (memory, orchestration, observability) that will capture the next wave of enterprise value.

I've been a founder, engineer, operator, and investor. That means I can help at every stage: from first product decisions to scaling GTM to navigating board dynamics. I've written code, closed deals, built teams, and run a P&L. I know what the problems feel like from the inside.

Read the full essay ↗

What I Look For

  1. AI technical depth + domain expertise Exceptional founders with strong founder-market fit over traction alone.
  2. Data moat quality & uniqueness As models commoditize, unique high-quality data with continuous feedback loops becomes the primary differentiator.
  3. Workflow embedding depth Products indispensable to users' daily work are defensible. Deep integration creates switching costs.
  4. Progressive defensibility Clear plan for layering-in defenses over time. Technical moats alone don't last.
Portfolio

Selected Investments

$550M+ deployed into AI across B Capital, Coatue, and angel investments. AI angel investments include Weights & Biases, Forethought, Braintrust, Haize Labs, Twin Labs, and Espa Labs.
Background

Operator-Turned-Investor in AI

Deep AI operating experience combined with a decade of early-stage investing.

  1. Google, Microsoft

    Engineering & Product

    I started my career in engineering and product at Google and Microsoft.

  2. Startups

    4× Founder & CEO

    Parsable (acquired by CAI), ChoiceVendor (acquired by LinkedIn), Mogad (acquired by iSkoot/Qualcomm), and Happiness Engines.

  3. Coatue

    General Partner, founding team of Coatue Ventures

    Built the enterprise investing practice for Coatue's Venture funds and led early AI investments including Weights & Biases, Covariant, Abacus.AI, Infinitus, Impira, and Raycast.

  4. Weights & Biases

    COO and CRO

    My deepest conviction bet. I angel invested at Seed (55x DPI), led the Series A & B for Coatue (10x blended DPI), and served as Board Director for 3 years. Then I left venture to join as COO, building global GTM from scratch and scaling the company from $3M to $50M ARR. As CRO, I hit or beat plan every quarter. Weights & Biases was acquired by CoreWeave in 2025.

  5. B Capital

    General Partner & Head of Tech Investing

    I led our investments in The OpenAI Deployment Company, Perplexity, Axiom, Goodfire, CodeMetal, Unblocked, and Collinear. I continue to invest behind the AI Co-Workers thesis.

Resources

Things I've Made

PDF · 6 pages

Negotiation Cheat Sheet

A 6-page pocket guide based on "Never Split the Difference." Thousands of people carry it into meetings.

Read more → Spreadsheet · Post ↗

Equity Valuation Tool

A spreadsheet and framework to help startup employees think through what their equity is actually worth.

Read the post → Reading list

Book Recommendations

What I'm reading and what's shaped my thinking, from Robert Caro to Carlo Rovelli to Frank Slootman.

See the list → How I work

Personal User Manual

If we work together, this is how to get the best out of me. Modeled on Urs Hölzle's approach at Google.

Read the manual → Since 2017

Meditation Resources

Practicing since 2017. Non-dual awareness, jhānas, and the books and teachers that have shaped my practice.

Explore → 6 Mac apps

Vibe-Coded Apps

Apps I vibe-coded with AI: Panda Contacts, Panda Expander, Panda Timekeeper, Switchyard, Granola Switcher, and SoundOrder. Downloads and install guides.

Get the apps →
Newsletter ↗

Substack

All posts on Substack →
Connect

Building AI co-workers? Let's talk.

I'm actively making new investments in AI co-workers (engineering, sales, finance, support) and the infrastructure that enables them (memory, orchestration, observability). $5M to $50M checks, from seed onwards.

The most useful first email is short: what you're building, where you are today, and what you're raising.

Reach out