Skip to content
Fund OS

Agents do the legwork, the team makes the calls

Autonomous agents handle sourcing through monitoring, so the team spends its time on value creation and the decisions that matter.

Tracked today
750+companies
Target: 5,000 in the funnel
Shortlisted today
200+
By AI screening of decks
Portfolio
18–22ventures
Final selection & partnership
Setup
Day one
Agent-to-agent and MCP
01How to play

An AI-native fund, by design

Agent-to-agent and MCP setup from day one. Autonomous agents handle sourcing through monitoring, so the team can focus on value creation and high-value decisions.

Deal funnel · target vs. today
5,000 tracked
Wide-funnel auto-sourcing from VCs, accelerators, universities, spin-outs · 750+ today
~1,250 shortlisted
AI screening of thousands of decks · 200+ today
~500 validated
Strategic matching to corporate needs
~100 engaged
Expert and AI-validated due diligence
18–22 in portfolio
Final selection & partnership

Bar widths on a log scale. Terracotta fill shows progress to date against target.

  • Autonomous deal sourcingFind top-tier climate-AI founders faster and earlier, scaling to thousands of inbound decks.
  • Founder decision engineBenchmark teams, tech and traction consistently, fast and bias-aware.
  • Dynamic investment structuringLive data shapes flexible, founder-aligned terms instead of boilerplates.
  • Embedded AI growth stackEvery portfolio team gets instant AI tools for go-to-market, ops and product.
  • Real-time intelligenceImpact, performance and competitors monitored live, so we act before anyone else.
02Running today

Not a roadmap: already in production

Live today, and used by the team every week.

Every week

Scout

Never sleeps

Agents scan the web for companies that fit our theses, read each one’s signature and score the fit. The strongest signals land on the team’s desk, one click from the CRM.

Every verdict

Learning loop

It learns from us

Every yes and no goes back into the scout, so next week’s shortlist looks more like what we actually back.

Every deal

Blind scoring

In each of our AI assistants

Independent views first, then the result. Each of us scores seven dimensions from our own AI assistant over MCP, and nobody sees a colleague’s score before submitting their own.

Built in-house

In-house stack

Our code, our data

Agents, pipelines and review tools in one codebase we own. New workflows ship in days, not procurement cycles.

03Operator-led

Agents free the time, operators go in

The hours the Fund OS saves go into the portfolio. Our operators are deployed to portfolio companies, hands-on, where they make the most difference.

Who

Operators, not only investors

Built and run companies

Our founding operating partner (founder of Greener, sold to CVC DIF) and operating Venture Partners who run energy and industrial companies today.

When

Where it moves the needle

Flagged by monitoring

Real-time portfolio monitoring shows where a company needs hands-on help, so operators go in early rather than at the next board meeting.

What

Inside the team

Peer to peer

Go-to-market, industrialisation, AI deployment, consortium building and the next round, working alongside CTOs, CSOs and CCOs as well as CEOs.

04Guardrails

Agents propose, the team decides

AI-first doesn’t mean handing over the keys.

  • People decideAgents rank and prepare. Every investment call is the team’s.
  • Everything on recordWho decided what, and why: our audit trail and the scout’s training data.
  • Our own bar firstThe responsible-AI standard we set for founders applies to us first.