From Hours to Outcomes: How AI Is Changing Enterprise Services

September 3, 2026

Interview with Lyoubomir Ovtcharov, Regional SVP Sales, Balkans, Adastra

You mainly work with large international organizations. What distinguishes Bulgarian companies when it comes to investing in data and AI?

Bulgarian organizations are highly pragmatic. They are open to innovation, but quickly focus on business impact, speed of implementation and operational efficiency. What is still sometimes underestimated is the foundation: clear data ownership, strong data quality and governance that makes information trusted, traceable and ready for AI.

Once value has been demonstrated, adoption can accelerate remarkably fast. For Adastra, this creates an opportunity not only to provide technology expertise, but also to bring practical experience from large-scale transformation programs – from Data and AI strategy and governance through implementation, adoption and managed operations.

That expectation of speed puts pressure on the traditional analytics model. What are clients asking for now that traditional analytics no longer delivers?

The market has moved well beyond traditional analytics. Organizations now expect data not only to explain what has happened, but to predict outcomes, optimize decisions in real time and increasingly automate execution. Technology is not the goal in itself. It creates value when it improves how an organization serves customers, manages risk or runs its operations.

At Adastra, this is driving an AI-first model in which strategy, trusted data, governance, cloud platforms and artificial intelligence work as one capability. We help clients move from ambition to production through AI strategy, AI-ready data, enterprise AI platforms, Generative and Agentic AI, process automation and AI-augmented delivery. The emphasis is on embedding intelligence into risk, compliance, customer engagement and operational workflows – with security, auditability and human oversight built in from the start.

A good example is a project for a major European banking group, where we unified customer communications and lead management across push notifications, in-app messaging, email, SMS, web channels and contact center workflows into a single orchestration platform. The result was faster campaign execution, a more consistent customer experience and lower operational effort.

If AI automates more of the delivery work, what happens to the way services are priced?

It changes fundamentally. As AI automates and augments more of the implementation and operational lifecycle, organizations increasingly expect scalable, outcome-oriented solutions rather than pricing based purely on human effort. They also expect existing delivery teams to use AI responsibly to increase throughput, quality and predictability. This is accelerating the shift toward platform-based services, AI-augmented delivery pods, accelerated fixed-price engagements and managed services, where value is measured by outputs and business outcomes rather than hours alone.

Does that mean the classic transformation project is disappearing?

No, but the classic project is becoming part of a broader lifecycle. Our commercial model is increasingly hybrid: transformation programs remain important, while long-term value is created through optimized platforms, reusable accelerators and managed services.

Organizations are moving toward continuous data and AI operations, where platforms, governance and AI capabilities evolve alongside the business rather than being delivered as one-off projects. This enables smoother adoption, stronger control and lower implementation risk.

AI-assisted delivery is also changing the full software and data development lifecycle. It can support requirements analysis, data discovery, architecture and mapping, code generation, testing, documentation, deployment and monitoring. The strongest model is not autonomous delivery, but expert teams working with governed AI tooling, quality gates and human review.

For example, after implementing a data platform, we typically continue supporting its operations, governance and development. AI-assisted engineering helps teams identify issues earlier, generate and test pipelines faster, keep documentation current and improve operational responsiveness – while accountability remains with experienced professionals.

Many organizations are still building AI use case by use case. Is that a mistake?

It is not necessarily a mistake; targeted use cases are often the right way to prove value and build confidence. The mistake is to stop there. Without an enterprise AI strategy, prioritization model, governance framework and reusable platform capabilities, isolated solutions become difficult to scale, integrate and control. We therefore help clients connect quick wins to an executable roadmap and a sustainable operating model.

Centralized, reusable enterprise AI platforms then provide common services for data access, model and agent lifecycle management, security, observability and cost control. They help organizations bring ideas into production faster while supporting regulatory compliance, risk management and internal policies. This creates a governed foundation for automation and AI-powered services across multiple functions.

What does that platform need underneath it to actually work in a regulated environment?

It starts with business priorities and an executable AI strategy – not with an architectural trend. In regulated environments, the roadmap must connect value, data readiness, ownership, risk appetite, regulatory obligations and the target operating model. Governance is not an annex added at the end; it is the mechanism that makes faster, safer scaling possible.

Technically, we see growing adoption of lakehouse and integrated enterprise data platforms, particularly where real-time analytics and AI are required. But architecture alone is not enough. Data quality, cataloguing, lineage, access controls, semantic consistency and clear accountability are essential for regulatory reporting, trusted decisions and responsible AI.

For one global financial institution, we replaced a legacy enterprise data warehouse with a modern lakehouse architecture. This enabled real-time credit scoring, automated data pipelines and a unified analytics platform while maintaining full governance and end-to-end data lineage.

Modern architectures are not built simply to store data. Combined with effective governance and strategy, they create trusted, reusable data products and enable AI at enterprise scale.

If value is measured in outcomes rather than hours, what do those outcomes look like in numbers?

The financial impact always depends on the use case. Automation and improved risk management can reduce costs by up to 30%, while better decision-making can improve operational efficiency by 10 to 20%. In financial services, even small improvements in data quality or decision speed can generate significant long-term value.

One example is an enterprise AI platform we delivered for a global bank. It reduced the time needed to deploy a new AI use case from months to weeks and generated more than $200,000 in savings during the first five months. This demonstrates both faster time to value and clear cost efficiency.

Another example comes from our banking migration projects. Our migration accelerator reduces implementation timelines by 50%, achieves 99.96% automated migration success and fully addresses data quality challenges during migration.

What ultimately decides whether a data or AI investment creates value?

Our role goes beyond implementing technology. We help organizations align Data and AI strategy with measurable business priorities, establish the governance and operating model, build the required platforms and use AI automation and AI-augmented delivery to move faster from idea to production.

Ultimately, value depends on three things working together: a clear strategy, trusted and governed data, and the ability to execute in complex real-world environments. That combination creates the confidence to scale AI responsibly – and turn investment into lasting outcomes.