Coforge is nearly doubling the size of its engineering centre in Kraków, taking it from more than 375 engineers today to over 660 professionals by 2027. The Indian IT services company said on 1 October that it had added more than 20,000 square feet of workspace at the site, which opened only last year, and will use it for what the company calls AI-first engineering, along with cloud and modernisation work. Its shares rose about 2% on the news.
The centre serves clients in travel, banking, logistics, retail and enterprise services, and Coforge is positioning it as a nearshore hub, close enough to European and UK clients to share working hours and allow frequent face-to-face contact.
"Europe is an important growth market for Coforge, and our continued investment reflects our long-term commitment to serving clients across the region," said John Speight, president and EU geography business leader. Chief operating officer Sunil Fernandes said the pace of growth reflected demand for "AI-enabled engineering and delivery models combining proximity, specialised talent, and speed."
What Coforge Is Actually Selling
The detail that distinguishes this from a routine office expansion sits in how the work is organised. Coforge delivers through what it calls Mod Squads, pod-based teams that combine AI agents with specialist full-time developers, and it sells them against business outcomes such as lower operating costs, shorter cycle times and better margins rather than against hours worked.
The traditional Indian IT model priced work by the number of people assigned to it. If AI agents handle a meaningful share of coding, testing and documentation, a provider can either keep billing by headcount and watch clients question the bill, or sell a smaller mixed team against results. Coforge has chosen the second route, and a nearshore centre with a few hundred senior engineers fits that shape better than a large offshore campus.
Fewer People, More Senior
A pod model changes who gets hired. The entry-level roles that once absorbed large graduate intakes, basic coding, manual testing and routine support, are the ones AI tools handle best, while the work that remains needs engineers who can direct agents, review their output and take responsibility for what ships.
Kraków's appeal lies there. Poland has a deep pool of experienced engineers, and a centre of 660 people in a European city costs considerably more per head than an offshore equivalent, which only makes sense if each person is doing higher-value work.
Why Poland, And Why Now
Kraków has become one of Europe's larger service delivery centres. The city had about 108,000 people working across 312 service centres by early 2025, according to the industry body ABSL, within a Polish sector employing 488,700 in dedicated centres and exporting $42.3 billion of services in 2024.
ABSL also found that AI had spread quickly through those centres, with cognitive tasks handled by AI at 90% of firms surveyed and more than 74% using intelligent automation to move staff onto higher-value work. For an Indian provider, that means Polish operations are already competing on the same terms rather than on cost alone.
European clients have their own reasons to prefer nearshore delivery, including data protection rules that make it easier to keep work inside the EU, and a preference for teams in compatible time zones when projects run as continuous engineering rather than fixed handovers.
Coforge Is Growing Fast Enough To Need The Space
The expansion comes from a position of strength. Coforge reported first-quarter revenue of $592.2 million for the three months to June, up 49% in rupee terms and 33% in dollars, helped by its acquisition of the engineering services firm Encora, which it said is now fully integrated. Profit after tax rose 110% in rupee terms, and its 12-month executable order book reached $2.23 billion, up 44% on a year earlier.
Chief executive Sudhir Singh said 86% of the company's revenues now come from AI-led engineering, data and cloud services. The company employs around 45,000 people, so Kraków will remain a small part of the whole, but a strategically placed one.
A Claim Worth Testing
That 86% figure deserves scepticism, as similar claims across the industry often count any project touching data or cloud platforms. Buyers evaluating providers on AI credentials should ask what share of delivery is actually performed by agents, what the provider guarantees about quality and security when agents write code, and how pricing changes as automation increases.
The harder question for clients is what happens to savings. If a pod delivers in half the time, a client wants the price to reflect that, while the provider wants to keep the margin. How that is settled in contracts will determine whether outcome-based pricing spreads or quietly reverts to headcount billing.
What It Signals For Technology Buyers
For European companies buying engineering services, the expansion is one more sign that the supplier market is reorganising around AI-assisted delivery and nearshore locations. Indian providers that built their businesses on offshore scale are adding European centres, and the pitch has shifted from cost per hour to speed and outcomes.
Buyers should expect more proposals built around small mixed teams, and should press for specifics: which tasks agents perform, who reviews their output, how client code and data are handled when agents touch them, and what happens if an agent introduces a defect. Those questions apply well beyond software projects, as AI agents take on project work across technology deployments.
Nearshore Europe Is Where The Services Model Is Being Rewritten
Coforge's Kraków plan is modest in isolation, a few hundred roles in a company of 45,000. Its interest lies in what it says about where the IT services industry thinks the next phase of work will be done, and by whom: closer to the client, with fewer and more senior people, alongside software that does a growing share of the building.
The test will come in the contracts. If clients can see what the agents do and pay for results rather than seats, the model holds. If not, Kraków becomes another delivery centre with a newer label, and the AI-first pitch will have to prove itself on something more measurable than headcount targets.
