Preview: London Tech Week 2026 — Who We Want to Interview & Key Companies to Watch

Key companies to watch—and why they matter

We're not tracking the loudest players.
We're tracking the ones making decisions that shape infrastructure.

Mistral AI

After its Series B, expectations are clear: move beyond models and into deployment.
The real question isn't performance.
It's positioning.

1. 01

   Can Mistral make open-weight models viable in enterprise environments?

2. 02

   How does it compete with API-first players on pricing without closing its stack?

3. 03

   What does edge inference actually look like in production—not in theory?

This is less about models, more about business model architecture.

Scaleway / Iliad Group

European cloud has a credibility gap—especially at the managed services layer.
Scaleway's expanding GPU cloud is one of the clearest attempts to close it.

What we'll be looking at:

1. 01

   Is the AI stack production-ready—or still infrastructure-first?

2. 02

   Can it support full pipelines without fallback to hyperscalers?

3. 03

   Where does it still lag behind in tooling and integration?

This is where sovereignty stops being a concept and becomes a product.

Darktrace

Cybersecurity is shifting toward AI-driven detection—but industrial environments raise the stakes.
Darktrace is pushing into OT/ICS systems, where:

• false positives are costly
• missed signals are catastrophic

We're not interested in demos. We want: deployment scale, measurable outcomes, failure cases.

Graphcore / SiPearl

Europe talks about sovereignty. Chips are where that conversation gets real.
Without control over silicon, everything else is downstream.

The question isn't whether alternatives to NVIDIA exist.
It's whether they're:

• production-ready
• energy-efficient
• integrated into real datacenter environments

This is the missing layer. And possibly the hardest to build.

Who we want to interview

We're not looking for polished narratives.
We're looking for informed trade-offs.

CTOs of European AI labs

Open-weight vs. closed models isn't ideology. It's a set of constraints: cost, control, performance, compliance.

We want to understand how those trade-offs play out in real deployments.

Infrastructure leads at sovereign cloud providers

Feature parity is the uncomfortable question.
Where do European clouds stand today—not in ambition, but in execution?
What still forces teams back to Amazon Web Services or Microsoft Azure?

CISOs and security architects

Security is no longer perimeter-based. We're looking at:

• Zero Trust adoption in regulated sectors
• post-quantum migration timelines
• real-world constraints, not frameworks

Datacenter operators

AI made infrastructure physical again. We want: real PUE numbers, cooling strategies under load, energy trade-offs at scale.

Technical themes we'll track closely

These are the signals beneath the announcements.

1. ▸

   AI inference at the edge

   Who is actually shipping models that run on-premise? Not demos—deployments.

2. ▸

   AI Act compliance tooling

   The first wave of compliance-as-a-service is emerging. The key question: Is it real infrastructure—or just regulatory wrapping?

3. ▸

   Confidential computing adoption

   TEE-based systems are moving beyond proof-of-concept. We're looking for: production use cases, performance trade-offs, integration complexity.

4. ▸

   European chip sovereignty

   Progress here will be incremental—but critical. Any reduction in dependency on external GPU supply chains changes the equation.

The underlying question

Across companies, interviews, and themes, one question keeps surfacing:

Not in parts.
Not in prototypes.
End-to-end.

London Tech Week won't answer that question.
But it will make one thing clear:
Who is actually trying—and who is still presenting.

For press accreditation or interview scheduling, reach us at press@europiatech.eu or visit our Press page.