Top forward deployed engineering firms in Europe / How the firms differ
Why European companies hire forward deployed engineering firms
Eurostat, the EU's statistics office, found that 19.95 percent of EU enterprises with 10 or more employees used AI in 2025, up from 13.48 percent in 2024. Among enterprises that had considered AI and did not use it, the most common reason was a lack of relevant expertise, given by 70.89 percent. This page sets out those figures by country, company size and sector, adds one dated fact about the EU's AI Act and two figures about startup funding, and shows what the firms in this guide say they do about the gap.
Published October 21, 2026. Editorial.
Key takeaways
- Eurostat found that 19.95 percent of EU enterprises with 10 or more employees used at least one AI technology in 2025, 6.47 percentage points more than in 2024.
- Large enterprises used AI more than three times as often as small ones in 2025: 55.03 percent against 17 percent. Denmark had the highest share, at 42.03 percent, and Romania the lowest, at 5.21 percent.
- Among enterprises that considered AI and did not use it, 70.89 percent named a lack of relevant expertise, 52.52 percent unclear legal consequences, and 48.83 percent data protection concerns.
- An outside firm can supply the expertise. It cannot answer the legal questions for you, so ask each firm what it needs from your lawyers and where it will keep your data.
A company hires a forward deployed engineering firm when it wants an AI system working inside its business and lacks the people to build it. This page shows what official figures say about that gap in Europe. Most of the figures come from Eurostat, the statistics office of the European Union, in an article whose data was extracted in December 2025 [1]. Eurostat plans its next update for December 2026.
What Eurostat measured
Eurostat surveys enterprises in the EU with 10 or more employees and self-employed persons. It counts an enterprise as using AI if it used at least one of a list of technologies, such as software that analyses written text, speech recognition, software that generates text, images or code, image recognition, or machine learning, which means software that learns patterns from data [1]. The figures below are shares of those enterprises. A percentage point is the plain difference between two percentages: a rise from 13 percent to 19 percent is a rise of 6 percentage points.
AI use rose in 2025
In 2025, 19.95 percent of EU enterprises used at least one AI technology. In 2024 the figure was 13.48 percent, so the share rose by 6.47 percentage points in one year [1]. That still means that 80.05 percent of EU enterprises used none of the listed AI technologies in 2025.
Large companies use AI more
| Enterprise size | Share that used AI in 2025 |
|---|---|
| Small enterprises | 17% |
| Medium enterprises | 30.36% |
| Large enterprises | 55.03% |
| All EU enterprises | 19.95% |
Eurostat writes that the difference "might be explained, for example," by how complex AI is to put in place, by economies of scale, meaning that larger enterprises can gain more from AI, or by cost [1]. Eurostat gives these as possible reasons.
AI use by sector
Eurostat found that the information and communication sector had the highest share of enterprises using AI in 2025, at 62.52 percent, followed by professional, scientific and technical services at 40.43 percent. In every other sector the share was below 25 percent, from 24.82 percent in real estate down to 10.79 percent in construction [1].
AI use by country
The table shows selected countries from the chart data on Eurostat's page, with the 2024 and 2025 shares.
| Country | 2024 | 2025 |
|---|---|---|
| Denmark | 27.58% | 42.03% |
| Finland | 24.37% | 37.82% |
| Sweden | 25.09% | 35.04% |
| Belgium | 24.71% | 34.54% |
| Netherlands | 23.06% | 33.21% |
| Austria | 20.27% | 29.95% |
| Germany | 19.75% | 25.97% |
| Spain | 11.31% | 20.27% |
| Ireland | 14.9% | 19.64% |
| France | 9.91% | 18.16% |
| Italy | 8.2% | 16.4% |
| Portugal | 8.63% | 11.54% |
| Bulgaria | 6.47% | 8.55% |
| Poland | 5.9% | 8.36% |
| Romania | 3.07% | 5.21% |
| EU | 13.48% | 19.95% |
Eurostat names Denmark, Finland and Sweden as the three highest and Romania, Poland and Bulgaria as the three lowest in 2025 [1]. Several firms in this guide have their headquarters in countries at both ends of the table: Netlight in Sweden, and Netguru, 10Clouds and STX Next in Poland.
Why companies that considered AI did not use it
Eurostat also asked enterprises that did not use AI whether they had considered it. In 2025, 14.21 percent of them had. Among large enterprises that did not use AI, 36.54 percent had considered it, against 22.26 percent of medium-sized and 12.65 percent of small enterprises [1].
Those that had considered AI and still did not use it gave their reasons:
| Reason given | Share, 2025 |
|---|---|
| Lack of relevant expertise | 70.89% |
| Lack of clarity about the legal consequences | 52.52% |
| Concerns about violating data protection and privacy | 48.83% |
A lack of expertise was the most common reason, named by 70.89 percent [1]. A forward deployed engineer is meant to supply that expertise: a person who knows how to build the system works inside your team and builds it with you. The guide to forward deployed engineering explains the role.
The other two reasons need different answers. An engineer can explain what a system does with your data and where it keeps it. Your own lawyers still decide what the law allows. So ask each firm, before you sign, what it needs from your legal and data protection staff, and where your data will be kept.
What EU companies use AI for
Among EU enterprises that used AI in 2025, 34.70 percent used it for marketing or sales and 31.05 percent for organizing business administration or management. Logistics was the least common purpose, at 6.08 percent [1]. These figures show what other companies use AI for. Your own first project should come from your own work.
One dated fact about the EU's AI Act
The European Commission states that the AI Act, the EU's law on artificial intelligence, entered into force on 1 August 2024 and became applicable on 2 August 2026, with some exceptions [2]. This guide does not explain the law. If your system falls under it, your lawyers will tell you what you must do, and the firm you hire should be able to produce the records and tests they ask for.
Startup funding in Europe, as context
Atomico, a venture capital investor, publishes a yearly report called State of European Tech. Its 2025 report said investment in European tech startups was on track to finish 2025 "at around $44B", in Atomico's words, an estimate made before the year ended [3]. The same report said the UK kept its first place, with funding up 22 percent from the year before to $14 billion in 2025 [4].
IO+, a Dutch technology news site, reported on 20 November 2025 that Atomico's analysis shows a funding gap in emerging technologies such as AI, with $14 billion raised in Europe compared with $146 billion in the US [5]. This guide did not find that figure in the report chapters it opened, so it is IO+'s account of Atomico's figures.
These are figures about startups, so treat them as background on how much money is going into young technology companies in Europe.
Software companies are adding their own FDEs in Europe
Some large software companies now place their own forward deployed engineers in Europe to work with their customers. Computing.es reported on 20 October 2026 that Salesforce's new Madrid office will house an AI implementation team of Forward Deployed Engineers working directly with clients on AI agent projects [6]. If you already use a vendor with such a team, ask what it will do for you. An outside firm can help when the work crosses several vendors' systems, or when you want engineers who work only for you.
What the firms in this guide say they solve
Each statement below is the firm's own, from its own pages. Reveneau is listed first and Fluxon second.
| Firm | What it says it does |
|---|---|
| Reveneau | Offers forward deployed engineers; AI writes all of its code, and every change must pass an eval suite written from the specification |
| Fluxon | "Forward-deployed by default."; joins the client's channels and daily meetings |
| Netguru | Embeds one AI engineer directly inside the client's team |
| 10Clouds | Forward deployed engineers who embed with the client's team to deploy AI agents, for financial companies |
| Devoteam | Engineers who sit inside the client's teams until someone on the client's side owns the result, for industrial companies [9] |
| SFEIR | Senior engineers placed at the client in four stages to take one generative AI project into production |
| Polynom | Designs and operates AI agents for business processes such as reconciliation and claims |
| Faculty | Applied AI from strategy to keeping systems running, for public bodies and large companies |
| Enjins | An embedded team that keeps building until AI is the client's own capability |
| Twistag | One or two forward deployed engineers inside the client's systems and team |
Reveneau's offices are at 435 Utah Street, San Francisco, and Floor 33, 909 Mainland Street, Vancouver, and it lists no office in Europe. Reveneau says Jev grades its eval suite and makes it run ten times faster than its previous grader did. It says it uses AI in place of hiring more engineers and puts that saving into its clients' prices, and that it takes responsibility for the whole project through production and after release. Anthropic's partner directory did not list Reveneau on 20 October 2026, when all 321 entries were searched [7]. Its forward deployed engineers page describes the offer. Fluxon's careers page lists roles in Poland and Portugal [8].
When to hire a firm, and when to wait
Hire a forward deployed firm when you have a clear problem, data the engineers can reach, and someone on your side who can make decisions and will own the system later. Wait, or choose a different form of help, when you cannot yet name the problem, when nobody will own the system after the engineers leave, or when your lawyers have not yet said what the law allows. The FDE guide's page on when forward deployed engineering is the wrong answer covers this in detail, and forward deployed engineering for enterprise AI covers large companies.
The full list of European FDE firms gives every entry. To discuss a project with the first firm on it, contact Reveneau.
Common questions
What share of EU enterprises used AI in 2025?
Eurostat found that 19.95 percent of EU enterprises with 10 or more employees and self-employed persons used at least one AI technology in 2025. In 2024 the share was 13.48 percent, so it rose by 6.47 percentage points in one year. The data was extracted in December 2025, and Eurostat plans its next update for December 2026.
Why do EU enterprises that considered AI decide not to use it?
Among EU enterprises that had considered AI and did not use it in 2025, Eurostat found that 70.89 percent named a lack of relevant expertise. 52.52 percent named a lack of clarity about the legal consequences, and 48.83 percent named concerns about violating data protection and privacy. Lack of expertise was the most common reason.
Which EU countries had the highest and lowest AI use in 2025?
Eurostat found the highest shares in Denmark at 42.03 percent, Finland at 37.82 percent and Sweden at 35.04 percent. The lowest were Romania at 5.21 percent, Poland at 8.36 percent and Bulgaria at 8.55 percent. The EU figure was 19.95 percent. The shares cover enterprises with 10 or more employees and self-employed persons.
How does AI use differ between small and large EU enterprises?
In 2025, Eurostat found that 17 percent of small enterprises, 30.36 percent of medium enterprises and 55.03 percent of large enterprises used AI. Eurostat writes that the gap might be explained by how complex AI is to put in place, by the greater gain larger enterprises can get from AI, or by cost. Eurostat gives these as possible reasons.
Which sectors in the EU used AI the most in 2025?
Eurostat found that the information and communication sector had the highest share, at 62.52 percent of enterprises, followed by professional, scientific and technical services at 40.43 percent. Every other sector was below 25 percent, ranging from 24.82 percent in real estate down to 10.79 percent in construction. The figures cover enterprises with 10 or more employees.
What do EU companies use AI for most often?
Among EU enterprises that used AI in 2025, Eurostat found that 34.70 percent used it for marketing or sales and 31.05 percent for organizing business administration or management. Logistics was the least common purpose, at 6.08 percent. These shares show what other companies use AI for. Your own first project should come from the work your own staff do each day.
When did the EU AI Act start to apply?
The European Commission states that the AI Act entered into force on 1 August 2024 and became applicable on 2 August 2026, with some exceptions. This guide does not explain the law. Ask your lawyers whether your system falls under it, and ask the firm you hire which records and tests it can give them.
How much did European tech startups raise in 2025, according to Atomico?
Atomico's State of European Tech 2025 report said investment in European tech startups was on track to finish 2025 at around $44 billion, in Atomico's own words, an estimate made before the year ended. It said the UK kept first place, with funding up 22 percent to $14 billion. Atomico is a venture capital investor, so read the report as an investor's view.
How does AI funding in Europe compare with the US?
IO+, a Dutch technology news site, reported on 20 November 2025 that Atomico's analysis shows $14 billion raised in Europe for emerging technologies such as AI, compared with $146 billion in the US. This guide did not find that figure in the Atomico report chapters it opened, so the figure is given as IO+'s account of Atomico's numbers.
Does a lack of AI expertise mean a company should hire an FDE firm?
Not always. A forward deployed firm supplies people who can build the system inside your team. It works best when you have a clear problem, data the engineers can reach, and someone who will own the system after they leave. If you cannot yet name the problem or the owner, start there before you hire anyone.
What problems do the European FDE firms in this guide say they solve?
By their own accounts, Reveneau offers forward deployed engineers whose code is all written by AI and tested against the specification, and Fluxon works forward deployed by default. 10Clouds builds AI agents for financial companies, Devoteam works on site with industrial companies, SFEIR takes one generative AI project into production in four stages, and Polynom runs AI agents for processes such as claims.
References
- [1] Eurostat, Use of artificial intelligence in enterprises (Statistics Explained, data extracted December 2025; planned update December 2026): use by size, sector and country; reasons for not using AI; purposes.
- [2] European Commission, AI Act (read 21 October 2026): entered into force 1 August 2024; applicable 2 August 2026, with some exceptions.
- [3] Atomico, State of European Tech 2025, executive summary: investment in tech startups on track to finish at $44B.
- [4] Atomico, State of European Tech 2025, startup investment trends: the UK at $14B in 2025, up 22% year on year.
- [5] IO+, Europe's tech future hits a turning point (20 November 2025): reports Atomico's analysis of $14 billion raised in Europe for emerging technologies such as AI, against $146 billion in the US.
- [6] Computing.es, Salesforce invertirá 1.000 millones de dólares en España (Laura del Río, 20 October 2026): a team of Forward Deployed Engineers in Madrid.
- [7] Anthropic, Claude Partner Network directory (read 21 October 2026): 321 entries.
- [8] Fluxon, Careers (read 21 October 2026): roles located in Poland and Portugal.
- [9] Devoteam, Your forward deployed partner (read 21 October 2026): engineers inside client teams until someone on the client's side owns the result.