Uniper cuts hiring delays with AI recruitment agent

Uniper has deployed an AI recruitment agent developed with Celonis and Microsoft Copilot Studio. According to the companies, it has cut the average time to sign candidates by 27 days.

The new setup has also reduced the average screening phase by 13 days, helping the energy company speed up hiring.

Uniper, one of Europe's largest energy groups, is using the agent through a natural-language interface that lets hiring managers ask questions about open job requisitions and shortlisted candidates. The system also flags urgent and overdue tasks for recruiters and managers, and can draft and send emails to colleagues and candidates.

The project centres on Celonis' Context Model, which gives the AI agent operational context rather than relying on raw data alone. Uniper and Celonis tested the system by running the same prompts twice: first using only data from Uniper's HR system, then using operational context from the model.

When asked, "What are the bottlenecks in the recruiting/hiring process?", the version using only HR data returned broad responses such as "data fragmentation", "hiring manager overload", and "missing task visibility". When linked to the context model, the agent produced answers tied to specific cases and users, including: "There are 27 candidates across various job requisitions who have been on the shortlist for more than 14 days."

Hiring process

That difference appears to have been central to Uniper's decision to move the system into full production. Recruiters and managers are now receiving more precise information instead of generic answers, along with clearer visibility into the recruitment process.

Robert Daamen, Head of Process Intelligence CoE at Uniper, outlined how the company sees the technology in its HR operations.

"With Celonis as our context layer, we're building AI solutions that aren't just tools we experiment with, but assistants we trust. The Celonis Context Model gives our HR teams complete transparency into the hiring and recruiting process and removes unnecessary hurdles that delays new talent from joining our team," said Robert Daamen, Head of Process Intelligence CoE at Uniper.

The work reflects a wider push by large companies to apply AI to internal processes where delays and fragmented information can slow decisions. Recruitment has become a common target because it involves multiple handovers between managers, recruiters, and candidates, as well as strict timing around interviews, approvals, and offers.

For Uniper, the project also shows how companies are trying to connect AI systems to live operational data so answers reflect current workloads and specific bottlenecks. In this case, the aim is not just to summarise information, but to direct staff to tasks that need immediate attention.

Broader rollout

The companies are also extending their work beyond HR into other parts of Uniper's business, including plant maintenance, hydropower operations, IT service management, energy sales and trading, financial services, and internal audit.

One example is sales and trading, where Uniper said its business executes hundreds of trades per hour. Celonis Data Core is being used to stream that data into Celonis in real time so teams can respond more quickly to operational developments.

The wider programme has backing from senior executives at Uniper, including the Chief Information Officer, Chief Financial Officer, and Chief People and Transformation Officer. The latter oversees HR and technology transformation, linking the recruitment work to broader operational change.

Lars Reinkemeyer, Chief Evangelist at Celonis, said the Uniper project shows how AI tools perform differently when they are connected to the way a business actually operates.

"Enterprise AI can only live up to its full potential when it understands how an enterprise actually runs and how to make it run better. The striking results Uniper achieved demonstrate how companies can generate tangible, compounding value by transforming business-critical operations with Enterprise AI grounded in context from Celonis," said Lars Reinkemeyer, Chief Evangelist at Celonis.