Morphism <> Moxxie VC — Meeting Brief

Date: Thursday, March 26, 2026, 11:00-11:25 AM PT Format: Zoom Duration: 25 minutes Attendees: Meshal Alawein + Alex Roetter (Moxxie Ventures) Intro by: Prateek Joshi (Moxxie) after productive call on Tue Mar 24 Coordinator: Jenn Miller (jenn@moxxie.vc)

Alex Roetter — Bio

Role: Managing Director & General Partner, Moxxie Ventures
Education: BS + MS in Computer Science, Stanford University
Twitter (2010-2016): SVP of Engineering. Led 2,000+ engineers. Built the ads network from near-zero to $2.5B/year in revenue. Scaled to hundreds of millions of users.
Kitty Hawk (Larry Page): President of the Flyer project. Oversaw 25,000+ crewed and uncrewed flights with 111 aircraft. Personally test-piloted after major software releases.
Angel investing: Coinbase, Veho, Spellbook, others.
Key trait: Engineer first, investor second. Expect detailed technical questions.

Moxxie Ventures — Overview

Founded: 2019 by Katie Jacobs Stanton (ex-Twitter VP, Obama White House, Google, Yahoo)
GP Team: Katie Stanton (Founder/GP), Alex Roetter (GP)
Fund III: $95M (closed July 2024)
Stage: Pre-seed and seed
Check size: $500K-$2M, sweet spot $1.5M at $15M post-money (~10% ownership)
Portfolio: 80+ companies, 3 unicorns. Notable: Coinbase, Carta, Calm, Lime, Opendoor, Clubhouse, Spellbook, Veho, Tipalti
Sectors: Enterprise software, infrastructure, vertical AI, climate, healthcare, fintech
Requirement: At least one technical founder, evidence of product beyond concept

Prateek Joshi — Your Internal Champion

Role: Investor at Moxxie Ventures (AI domain specialist)
Background: Founded Plutoshift (AI for industrial automation). Career started at Nvidia. Author of 13 books on ML/AI. Hosts "Infinite Curiosity" podcast (140+ countries).
Signal: His "fruitful chat" on Mar 24 + intro to Alex means the AI governance thesis passed his technical filter.

25-Minute Agenda

| Time | Block | |------|-------| | 0-5 min | Rapport. Thank Prateek for the intro. Quick personal connection. | | 5-13 min | Problem + Solution. Lead with the pain, not the math. | | 13-18 min | Demo/walkthrough (CLI output, proof witnesses, MCP integration) | | 18-23 min | Q&A | | 23-25 min | Next steps. Offer a longer follow-up. |

Morphism Pitch (for this meeting)

Problem (2 min)

Engineering teams are adopting AI agents (Claude Code, Cursor, Copilot) with zero governance layer. Every agent interaction introduces entropy: config drift, policy violations, inconsistent behavior across repos. Current solutions are observability-only — logs after the fact. There is no "verify-then-execute" pipeline for AI agent behavior.

As teams scale agent usage, "did this AI change break our invariants?" becomes unanswerable without a formal framework.

Solution (3 min)

Morphism is the control plane for AI-assisted engineering. A mathematically rigorous governance framework using category theory to model and verify agent behavior.

Four-stage pipeline:

READ — Load repository state
VERIFY — Check invariants against governance policies
EXECUTE — Apply transformation
EMIT — Produce audit trail with Proof Witnesses (tamper-evident, HMAC-SHA256)

Zero-dependency, offline-first core. Runs locally or in CI. No vendor lock-in.

Product Surface (2 min)

| Tier | What | Price | |------|------|-------| | Free | CLI (morphism init/validate/score/doctor) | $0 | | Pro | MCP Server for Claude Code/Cursor + SDK | $29/seat/mo | | Team | SaaS dashboard for governance visibility | $79/seat/mo | | Enterprise | Unlimited agents, custom policy rollout | Custom |

Traction (be honest)

7 npm packages published under @morphism-systems/
Dual-language implementation: TypeScript packages + Python category theory engine
Pricing tiers defined and gated in code
Active development: proof compression, CLI dashboard, tamper-evident signatures
No paying customers yet — be upfront about this

5 Questions to Ask Alex

"What patterns are you seeing in how portfolio companies manage AI agent adoption across engineering teams?" → Opens with their experience. Positions Morphism as the answer to pain they're already seeing.
"You scaled Twitter engineering to $2.5B in revenue with 2,000+ engineers. What governance tooling did you wish existed at that scale?" → Personal, shows homework. Let him describe the problem in his own words.
"Moxxie invested in Spellbook (AI contract copilot). How do you think about the governance layer for AI-assisted development specifically?" → Validates whether they see governance as an investable category.
"Our core thesis: AI governance needs mathematical rigor, not policy documents. Category theory gives composable, verifiable primitives. Does that resonate?" → Tests if technical depth is a feature or concern. Alex's Stanford CS background can engage with this.
"What would you need to see from Morphism in 3-6 months to feel confident about the investment thesis?" → Direct, respects the short call, gives a clear roadmap for follow-up.

What to Avoid

Don't oversell traction. Be precise about shipped vs. in-progress. Don't imply paying customers.
Don't lead with the math. Problem + business case first. Category theory is the differentiator, not the opener.
Don't badmouth existing tools. Position Morphism as complementary ("governance layer for AI agents"), not competitive.
Don't go over time. 25 minutes is short. Offer a longer follow-up if needed.
Don't forget Prateek. "Prateek and I had a great conversation about X" reinforces the warm intro.
Be ready for "who's on the team?" Moxxie requires at least one technical founder.
Don't signal blockchain. Tamper-evident proofs use HMAC, not blockchain. Moxxie explicitly excludes blockchain investments.

Sources

Moxxie Ventures: moxxie.vc
Moxxie Fund III: TechCrunch, July 2024
Alex Roetter: Crunchbase, Signal NFX, Waydev interview
Prateek Joshi: Moxxie Substack announcement
Katie Stanton: Wikipedia