← About · Focus areas · AI & deeptech
Focus areas Talent Go-to-market

Discover. Invent. Scale.

AI and deeptech offer outsized leverage on our biggest economic and societal challenges. We back founders building in the Netherlands and Europe to seize them — to discover, invent and scale to transformative impact from here.

01 — Position in brief

AI is rewriting the playbooks faster than anyone can learn them — raising the frequency of one-way-door decisions even as the opportunity for impact has never been higher. We believe founder decision quality is the key driver of the ability to scale, and that social capital, the right ecosystem conditions, and culture are what raise it. Our focus is backing founders building in the Netherlands and Europe to invent at the frontier and scale to transformative impact from here.

02 — How we think about our role

Three lenses on the work.

Which kind of AI or deeptech company we're backing, and where we can help most: Deeptech × AI, Frontier AI, and Transformative AI. Invention at one end, impact at scale at the other — and our job is to connect them.

Discover
AI that compresses discovery — faster simulations, self-driving labs, new materials and molecules.
What it is

Companies using AI to push the frontier of other domains — physics, chemistry, biology, materials. AI is changing how science itself gets done: faster iterations, faster simulations, self-driving labs that compress years into months. Many of the breakthroughs a decarbonised economy needs — greener steel, cement and plastics, recovered metals, entirely new materials — increasingly start here.

The hard part

Deeptech is never fully lean: built from many components, you get only a handful of iteration cycles before manufacturing, and financing first-of-a-kind builds or shaping markets that don't yet exist is something the VC playbook alone can't do. The frontier keeps moving, too — progress in AI research compounds what's possible next. Make each cycle count.

Invent
New architectures, and the silicon and energy that make AI possible.
What it is

Two strands. AI research — new architectures for the hardest open problems, like continual learning (Yann LeCun's AMI programme is the archetype). And AI hardware and the stack it runs on — chips, photonics, energy-efficient and demand-flexible compute — because today's compute-and-power curve isn't sustainable, and this is where Dutch deeptech has structural strengths (ASML and its adjacencies).

The hard part

Energy efficiency is now the binding constraint — power demand keeps rising despite efficiency gains, and Europe's electrons are scarcer and pricier than the US's or China's. For a congested, space-scarce Netherlands that pressure is as much opportunity as obstacle: the way out runs through hardware-level breakthroughs, not just code.

Scale
Not technical milestones — economic transformation, at industrial-revolution speed.
What it is

Defined by outcomes, not benchmarks — “a transition comparable to, or more significant than, the agricultural or industrial revolutions, but over a much shorter period.” The useful test is the economic Turing test: could you tell whether a series of tasks was done by a human or a machine? These are the AI-native stacks remaking accounting, legal, customer support, software development and healthcare.

The hard part

Distribution, category-creation and global reach decide who wins — not the cleverness of the model. The prize is replacing incumbents, not upgrading them. (Corgi, in US insurance, is an early example of the new wave.)

03 — What this looks like in practice

Three ways the community compounds your AI edge.

Concrete sessions and access — grounded in the founders and researchers we already bring into the room.

01

Learn from peers

Founder-to-founder sessions on putting AI to work today — GTM, operations, governance, token spend, and building software with agents — where the sharpest playbook comes from a peer who cracked it last quarter, not a slide. Recent examples: ElevenLabs' Luke Harries on repeatable growth systems, Encord's AI sales engine, and a working session on using AI agents to clear engineering bottlenecks.

02

Europe's best

The fastest-scaling companies in Europe have already worked out how to leverage AI — so we put founders in the room with them. Recent gatherings have convened ElevenLabs, Encord and Ineffable Intelligence, alongside Synthesia, Miro, Hugging Face and Weaviate — sharing candidly what's worked and what hasn't.

03

From the frontier

Direct access to the people defining what comes next — world-leading researchers like Max Welling (CuspAI), and teams operating where AI meets the physical world, like Callosum. The long-range understanding you can't get from the headlines.

Proof, not promises — a sample of the rooms we've already put founders in
San Francisco · Oct 2025 Thirteen founders inside the AI build-out — Salesforce, OpenAI, Anthropic, General Catalyst. Read the recap → London · Jun 2026 Eight founders, three closed-door rooms — Salesforce, Google and the UK's £500m Sovereign AI Fund. Read the recap → Amsterdam · Dec 2025 Europe's top AI founders — Synthesia, Miro, Hugging Face, Weaviate, n8n — five roundtables, no keynotes. Read the recap → New York · Mar 2025 Thirteen CEOs in rooms you can't book — Adyen, Insight, Wilson Sonsini, NASDAQ to close. Read the recap →
Invent here, scale from here

Invent at the frontier,
scale to transformative impact
from the Netherlands.

BR
Bob Rietveld
Managing Director · Techleap

Building at the AI or deeptech frontier — or have a thesis you think we should back?

Get in touch · bob@techleap.community →