EPAM accelerates agentic AI adoption for enterprises

Partner of the Year Award winner EPAM teamed up with Albert Heijn to deploy a secure AI assistant, simplifying retail workflows and setting up enterprise-wide transformation.

January 30, 2026

What does it take to move from AI experimentation to enterprise-wide scale? For EPAM, the answer emerged from a year of foundational work—helping customers fix data estates, aligning stakeholders, and building repeatable engineering practices. As a Microsoft Global Systems Integrator (GSI), EPAM supports customers preparing for scalable adoption, with a focus on agentic AI and production-grade deployments.

The company’s long-standing partnership with Microsoft includes 17 specializations across AI, app innovation, and data, as well as designations across all Microsoft solution areas. And the company is invested in internal enablement, deploying Microsoft 365 Copilot to more than 2,000 employees and reducing external collaboration hours by 20%. Together, their well-established credentials and internal enablement efforts position EPAM to deliver high-value engagements and accelerate innovation for enterprise customers.

This expertise, partnership, and proven ability to deliver value is part of what earned EPAM the 2025 Innovate with Azure AI Platform Partner of the Year Award. Their partnership with Albert Heijn, a leading grocery retailer in the Netherlands, was an example of this award-winning work. That deployment was just one part of a broader posture as the team focuses on helping customers reimagine business processes with AI—moving beyond pilots and into production.

A year of experimentation, not scale

For many organizations, 2025 was a year of experimentation with generative AI—but not yet scale. EPAM saw customers across industries testing new capabilities, often in isolated pilots. “A lot of companies tried,” said Dmitry Tikhomirov, Vice President, Technology Solutions at EPAM. “Some of them failed, and very few scaled generative AI to production.”

EPAM identified two key barriers to broader adoption. The first was data readiness. Many organizations lacked the infrastructure to support enterprise-grade AI. “It’s very hard to scale AI if your data is not ready for that,” Tikhomirov explained. “Building out data platforms and moving your data closer to AI models—modernizing and simplifying data—was a big trend.”

The second challenge was organizational alignment. In many cases, AI initiatives were driven by technical teams without full buy-in from business stakeholders. “If you just patch part of a process with AI, you’re not necessarily accelerating the entire process,” said Tikhomirov. “Your return on investment is less tangible, in such cases.”

EPAM spent the year helping customers address these foundational gaps. That included fixing data estates, navigating compliance, and aligning internal teams around shared goals.

A person seated using a laptop computer.

It’s very hard to scale AI if your data is not ready for that. Building out data platforms and moving your data closer to AI models—modernizing and simplifying data—was a big trend.

—Dmitry Tikhomirov, Vice President, Technology Solutions, EPAM

Operationalizing enterprise agentic AI in retail

EPAM partnered with Albert Heijn to deliver a production-grade generative AI platform using Microsoft Azure AI Foundry. The solution embedded an employee-facing virtual assistant directly into the retailer’s staff app, designed to simplify restocking, accelerate onboarding, and improve access to product and inventory information. The assistant can orchestrate multi-turn interactions, retrieve authoritative data, and automate task flows—all under enterprise-grade governance and observability controls.

Graphic showing success rate of Copilot internal deployment.

The architecture leveraged Azure OpenAI to deploy a virtual assistant and was hosted on Azure Kubernetes Service, a solution that could deploy and scale the app. Using Azure OpenAI, the conversational assistant can answer a range of questions from what needs to be restocked on shelves to where a product can be found in store. The company’s data was stored using Azure Database for PostgreSQL, a fully managed database that can integrate with AI tools at a low cost.

EPAM frames the project as an example of enterprise agentic systems: assistants that not only respond to queries but also execute back-end API calls and log actions for auditability.

Bottom line is what we can do with AI to make ourselves faster .... Top line is, how can you change a business to embrace AI starting from business processes?

—Dmitry Tikhomirov, Vice President, Technology Solutions, EPAM

Building momentum for enterprise-wide AI

EPAM sees 2026 as an inflection point. After a year of foundational work—data estate fixes, compliance reviews, and internal alignment—customers are now positioned to move forward with confidence. “Capabilities are there,” Tikhomirov said. “They passed through complex compliance and security discussions. This year is going to scale.”

The team is already seeing repeatable outcomes in software engineering acceleration and a shift in mindset among business leaders. “We see more and more desire and understanding on the business side about how they can reinvent the business model with AI,” Tikhomirov said.

EPAM plans to expand their learnings from this deployment to include broader AI offerings and showcase how their posture is evolving—and how they can help customers do the same. Tikhomirov noted a confluence of AI across engineering and business. “Engineering AI and business AI are converging on the same agentic platform,” he said. “Organizations have started to build two or three platforms, but consolidation at the platform level plays quite well for us.”

Explore more Partner Success stories

Discover how organizations like yours are using Microsoft technology to help customers solve challenges, drive results, and scale their businesses.

Browse all stories

This document is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.