Ask any founder or CTO what their hiring plan looks like, and you’ll get a different answer depending on which round they last closed. That’s not a coincidence. Across pre-seed to IPO, the balance between contract and permanent talent shifts in a fairly predictable pattern. For anyone recruiting AI, data, software and product talent across Europe, understanding that pattern is the difference between placing candidates reactively and actually advising clients on workforce strategy.

Having recruited across this lifecycle, and now spending most of my time in critical infrastructure organisations, I’ve watched the same story play out again and again. Here’s how it tends to go, and why the critical infrastructure sector is, in many ways, the exception that proves the rule.

Pre-seed and stealth: why contractors make sense first

In pre-seed and stealth mode, headcount is a liability before it’s an asset. Cap tables are tight, product market fit is unproven, and the founding group is often five to fifteen people who need to move fast without the overhead of a large permanent structure. This is where specialist contractors do some of their most valuable work. A fractional ML engineer can get a proof of concept into production. A contract data engineer can build the first pipeline. An interim CTO can shape the architecture before a permanent hire is even affordable.

Isomorphic Labs is a useful example of the mindset, even though it emerged from Google DeepMind rather than a garage. It spent years operating quietly on AI driven drug design before going public with its ambitions, then raised $600 million in its first external round in March 2025 and a further $2.1 billion Series B in May 2026. That pattern of years under wraps followed by a sudden scale up is exactly what stealth AI ventures across London, Paris and Berlin are trying to replicate, and it’s precisely the phase where lean, contract heavy delivery makes sense, because nobody yet knows which roles need to be permanent.

The goal at this stage is almost never to build a large team. It’s to reach Series A with proof that the product works, using just enough contract support to get there.

Series A: the pivot to permanence

Series A is where the story usually changes. This is the point where organisations stop optimising purely for speed and start optimising for ownership. Founders want people who will still be there in three years, who think like owners because they legally are, through equity packages and meaningful titles like “founding engineer” or “founding data scientist.” Mistral AI’s growth is a good illustration of scale at this stage. Even past Series A, it has run more than 160 open roles across Europe, heavily concentrated in Paris and London, almost entirely on a permanent, in office or hybrid basis. The message to candidates is consistent: come and help build the core, and be rewarded with a stake in what you build.

This is also where AI, data, software and product hiring gets genuinely competitive, because every well funded Series A company in Europe is chasing the same small pool of senior ML engineers, platform engineers and product leads who can operate with little support. Contract use doesn’t disappear here, but it narrows to genuinely temporary needs: a compliance specialist for a specific market entry, a UX researcher for a discrete project, a contract recruiter (fittingly) to help build that first permanent team at speed.

Series B through D: build the core, keep contractors for the edges

From Series B onward, the pattern from Series A largely holds, just at greater scale. Organisations are building out functional depth, a real data platform team, a proper MLOps function, a product organisation with multiple pods, and they’re doing it with permanent hires, equity, and increasingly structured career paths. Helsing, the German defence AI company, shows how fast this can move. A €600 million Series D in June 2025 valued the business at roughly €12 billion, and by May 2026 it was in advanced talks for a further $1.2 billion at an $18 billion valuation. Quantexa followed a similar arc in the UK data and AI space, closing a $175 million Series F in March 2025 at a $2.6 billion valuation, growing revenue nearly 50% year on year, and openly discussing a 2026 IPO.

Companies at this stage still reach for contractors, but typically for niche, time boxed work rather than core capability. Think a specific data migration, a security audit ahead of a major public sector contract, or a burst of extra engineering capacity around a product launch or regulatory deadline. The core philosophy remains build permanent, contract at the margins, because the equity story is still compelling and there’s still a clear “we’re going somewhere” narrative to sell top permanent talent on.

Public markets: where the equity story runs out

Then comes the IPO, or the equivalent liquidity event, and the calculus flips. The tool that made permanent hiring so persuasive, meaningful, potentially life changing equity upside, is largely gone. Shares are liquid, often immediately sellable, frequently underwater relative to private valuations, and no longer feel like a lottery ticket. Public companies also face new pressures: quarterly reporting, cost discipline, governance, and shareholders who don’t care about “team building” narratives. The people who joined at Series A for the equity dream often leave once that dream has been priced by the public market, sometimes for the next stealth startup, restarting the whole cycle.

Klarna’s experience since its September 2025 NYSE listing is instructive, if not a perfect match for the pattern. In the run up to IPO, Klarna cut around 700 roles and leaned hard on AI to replace them. Shortly after listing, CEO Sebastian Siemiatkowski publicly admitted the company had “gone too far,” acknowledging that AI couldn’t fully substitute for experienced people, and Klarna began rehiring. Whatever the specifics, it’s a clean example of a public company discovering, in real time, that cutting permanent headcount without a plan for specialist delivery capacity creates a capability gap, one that contract talent is usually the fastest way to close.

Darktrace took a different exit from the same pressure. The Cambridge founded cybersecurity AI business listed in London in 2021, then was taken private by Thoma Bravo in a $5.3 billion deal completed in October 2024, a reminder that going public isn’t always the final stage of this lifecycle, and that public market scrutiny is itself a reason some businesses choose to step back out of it.

The broader point holds regardless of which path a company takes. Once equity stops being the main draw, and delivery obligations don’t slow down, contract specialists become the practical way to access senior AI, data and engineering talent without competing on an equity story you can no longer tell.

Critical infrastructure: permanently in the “post-IPO” position

This is where critical infrastructure organisations are a special case worth calling out on their own, because most of them never had the equity story to begin with. Energy operators, water utilities, telecoms, rail and grid operators are, from a talent economics perspective, permanently in the position a newly public tech company finds itself in. No meaningful equity upside, heavy regulatory and governance overhead, and an absolute requirement to keep delivering regardless of market conditions.

What’s changed is the scale of the digital and AI demand these organisations now carry. UK data centres have effectively been classified as critical national infrastructure, sitting alongside energy and water in planning terms, and the National Energy System Operator (formerly National Grid ESO) is running digital transformation programmes, including digital twins and net zero modelling, that require the same calibre of data engineering and AI talent as any venture backed scale up. The difference is that a water company or grid operator can’t offer a founding engineer package or a pre-IPO options grant to compete for that talent. Contract engagement, at senior day rates, is often the only realistic way to bring in genuinely scarce skills for a defined transformation programme.

That makes critical infrastructure a contractor heavy market for AI, data, software and product specialists, not because these organisations are going through a funding stage in the venture sense, but because they sit permanently in the part of the lifecycle where equity can’t do the talking.

What this means in practice

For anyone hiring or recruiting into this space, the practical read is straightforward. Pre-seed and stealth clients want contractors who can move fast and leave a clean handover. Series A to D clients want permanent people who’ll buy into an equity story, with contractors reserved for genuinely temporary needs. Post-IPO and large cap clients, and critical infrastructure organisations by default, need a different pitch entirely: access to senior, proven AI, data, software and product talent who aren’t chasing an equity outcome, delivered fast enough to matter and priced to reflect genuine scarcity rather than promise.

Understanding which stage a client is actually in, not which stage they think they’re in, is what separates a recruiter who fills a role from one who’s trusted to shape the hiring strategy.

About the author

If you are a technology contractor looking to work on high-impact programmes in critical sectors, or an organisation seeking proven specialists who can deliver from day one, connect with Cam Dalziel at Aspire Technology or email Cameron.Dalziel@aspirerecruitmentgroup.com.

Cam works closely with contractors and organisations across global technology markets, helping build the teams that deliver secure, resilient, and future-focused outcomes.