Axiom
Contact

AI-only investing

We back AI startups exclusively

Axiom gives LPs, founders, and talent a fast read on the firm's thesis, current conviction, and deal-level detail.

View portfolioRead insights

Current lens

Selective capital for the AI stack
The firm stays concentrated on infrastructure, security, robotics, and workflow applications where adoption compounds.
FocusAI-native software
StrategyLead or co-lead
LensOperator-backed diligence

The problem

Generalist capital misses what matters in AI.

Too many firms treat AI as a theme. Axiom treats it as the whole market structure and follows the companies that can survive each shift in the stack.

Less noise, more pattern recognition.
Faster diligence across similar technical cycles.
Better follow-on conviction when the market turns.

How we invest

A narrow filter, applied consistently.

AI-native

The product only makes sense if intelligence is central to the workflow.

Adoption-led

The first use case must land in a real habit, not a demo loop.

Production-minded

Reliability, governance, and data hygiene are treated as product features.

Proof

01

The portfolio is the thesis.

Axiom keeps a close view of where the stack is moving and backs a small number of companies that map to that shift.

Featured position

Nova Lattice
A control plane for teams shipping multiple models into regulated workflows.

Key metric

3x faster model rollout

Detail

12 regulated deployments

Detail

$0.8m in annualized contract value

Active positions

AI infra, security, and workflow
Signal HarborSeed

Policy and access control for prompts, tools, and agents.

Atlas MotionSeries A

Vision-language planning for warehouse and factory robots.

Hearth BiomapPre-seed

Clinical note synthesis and retrieval for specialty care teams.

See the full portfolio

Current conviction

The stack changes fast. Our filter does not.

Axiom keeps the portfolio concentrated in the places where AI creates durable product advantage: production tooling, trust layers, robotics, and daily workflow software.

Back founders with real operating context.
Stay close to the customer workflow.
Prefer products that become infrastructure.

Team

Operators with a clear point of view.

The partners spent years building products for technical buyers before backing companies that need the same level of judgment.

Managing partner

Maya Chen
Maya backs founders who can make AI feel operational, not theatrical. She spent a decade building products for technical buyers before moving into venture.

Prior wins

  • Led product at a category-defining infra company
  • Advised two AI teams through their first enterprise sales cycles
  • Built a seed fund thesis around workflow software

Credibility markers

Keynote speaker, Applied AI ForumFormer board observer, enterprise ML companyPublished essays on model governance and adoption

Partner

Daniel Okafor
Daniel comes from the operator side of machine learning systems and has spent his career turning messy technical constraints into product advantage.

Data, evaluation, observability, and production readiness.

Guest lecturer, systems design for AI teamsJudge, university startup showcaseContributor to open technical communities

Partner

Priya Shah
Priya has a product instinct for AI experiences that land with real users. She spends most of her time with founders close to distribution and workflow change.

Agents, vertical AI, and the path to repeatable adoption.

Featured speaker at operator dinnersMentor, early-stage founder programsFormer reviewer for an AI accelerator

Insights

Memo-style writing for founders and LPs.

Short, direct notes on the shifts that matter to AI startups, from diligence to distribution to infrastructure.

Thesis6 min read
Why AI-only portfolios compound faster
Concentration is not a brand choice. It changes the quality of the pattern recognition you build.

A focused portfolio sees the same technical and market signals again and again, which sharpens conviction more quickly than a generalist strategy can.

January 2026

Open memo
Diligence5 min read
What we look for before model quality
The best signals show up before the benchmark curves get impressive.

Most investors overfit to a demo. We want to know whether the product already fits the user's day and the buyer's budget.

December 2025

Open memo
Infrastructure7 min read
Why infrastructure still compounds
AI infra is not a commodity when the workflows are regulated, production-heavy, or expensive to get wrong.

The right infra company is not just faster. It removes enough friction that the buyer can scale the system with confidence.

November 2025

Open memo
Strategy4 min read
When to lead in AI applications
Lead when the product already earns its place in a daily workflow and the team can explain the next expansion path.

Leading a round is easier when the company has already earned pull from the people closest to the work.

October 2025

Open memo

Start here

If you are building the AI stack, we should talk.

Founders, LPs, and operators can start a conversation with the team or send a direct introduction for a specific company, round, or thesis.

Contact Axiom