The agents are coming: How Agentic AI is transforming the financial world
News, 03/12/2025
At Structured Finance 2025, Sebastian Denef, PhD, CEO & Co-Founder of AGENTS.inc, Thomas Heinz, CGO Sopra Financial Technology, and Martin Stolberg, Head of AI, Sopra Steria, presented what is arguably one of the most important ideas for the future of the financial industry. His presentation, “The Agents Are Coming,” makes it clear:
The era of agentic AI has begun—and it will change everything we know about data, decision-making, and efficiency in the financial sector.
From “Good morning” to M&A: Why agents are becoming the new coworkers
The introduction is deliberately humorous: using everyday examples such as “Dinner's ready” or the question “What is my next M&A target?”, Denef shows how we already live with assistant systems today – but how limited classic co-pilots still are in a corporate context.
The central thesis:
We are on the threshold of autonomous, professional agents that no longer just respond, but act, analyze, warn, and prepare decisions independently.
Source: AGENTS.inc
Why today's AI is not yet sufficient
The picture from the Gartner Symposium underscores an important statement: AI will not be ready for real decision-making authority by 2025—at least not in the form that many companies would like to see.
But technologies are developing rapidly. And this is exactly where AGENTS.inc comes in.
Introducing AGENTS HQ – the AI agents are coming
The core of the presentation is the introduction of AI agents, which form the systemic basis for professional AI workforces.
- The Coordination Hub: A proprietary system that distributes complex workflows across thousands of specialized agents.
- The Trust Layer: A shared database that links internal and external data and prevents hallucinations system-wide.
This architecture enables a new form of collaboration between humans and machines.
Three stages of agent development
Sebastian Denef structures agents according to three levels of maturity:
Level 1 – ANALYZE
Agents analyze data, generate reports, and deliver executive insights with
✔ 100% accuracy
✔ 0% hallucination
✔ >1000x data compression
Level 2 – WARN
The system detects risks early on, prioritizes information, and warns of developments that human teams alone would hardly be able to grasp at this speed.
Level 3 – ACT
The most advanced step: agents become capable of acting. They carry out defined actions independently – securely, transparently, and verifiably.
The impressive ROI figures: speed, costs, coverage
The core of the argument: Agentic AI delivers measurable, economically relevant results.
Examples from real-world applications:
Consulting processes
• 4 weeks of human work → 1 hour with agents
• 672x faster, 200x cheaper
Supplier search
• Monitoring 10 cities → 400 cities possible
• 20x more coverage, 66x cheaper
Regulatory monitoring
• $100,000/year → $4,000/year
• 25x cheaper, higher precision
Reputation management
Agents understand contexts rather than just keywords. These figures are not theoretical – they come from real enterprise implementations.
What companies misunderstand: The 80/20 rule does not apply to AI
One of the most important slides illustrates that the usual 80/20 rule fails in AI projects.
Why? Because the visible 20% prototype “showpiece” says nothing about the hard infrastructure underneath – the crucial 80% that delivers real value (data management, governance, model selection, scaling, error control).
AI agents are only 10% AI – and 90% infrastructure
Denef sums it up clearly: Without robust data pipelines, trust layers, internal data connectivity, and orchestrated multi-agent systems, no AI can scale.
This is an important message for banks, insurance companies, and corporations that are testing AI use cases – but not (yet) putting them into production.
Practical examples: Where agents are already working today
The presentation showcases use cases such as:
- Credit and financial risk analysis in global credit portfolios
- Reputation analyses
- Market data screening
- Supplier identification
- Compliance and regulatory monitoring
Many of the world's leading enterprises are already using agents productively.
The message at the end: YES. WE CAN.
Sebastian Denef concludes with a clear statement: “YES. WE CAN.”
Agents are no longer a vision of the future. They are reality – faster, more precise, and more scalable than ever before.
His invitation: “Let's connect, let's talk.”
Conclusion: Agentic AI is the next big leap – and the financial industry is at the center of it
The presentation impressively shows that AI is just the beginning. Agents are the next productivity boost. Companies that invest now will gain a sustainable competitive advantage.
The future of the financial industry will be shaped by autonomous agents – and it starts now.
At Sopra Steria, we support the operation of AI agents with our Sopra Steria Trusted AI Platform.
Have we sparked your interest?