AI is changing how software gets built, but in regulated industries speed without control quickly becomes a liability. This session offers a practical view on using AI across the SDLC in banking and other sensitive domains: what to allow, what to gate, and how to keep humans accountable without slowing teams down. Drawing on hands-on experience delivering digital banking solutions for banks and responsible AI governance, it covers data boundaries, AI-generated code, and governance built into the delivery process rather than added on later.

Autonomous agents represent the next evolution of business automation. Instead of simply answering questions, these agents can reason, make decisions, interact with business systems, and execute processes on behalf of users. But how much autonomy should an agent have? When should it escalate to a human?

In this session, we will explore how Microsoft Copilot Studio enables organizations to build autonomous agents that can monitor events, make decisions, trigger actions, and collaborate with people across Microsoft 365. We will discuss best practices for designing agent behaviors, implementing guardrails, managing approvals, and ensuring responsible AI governance.

Expect a live demonstration showing how to build and configure an autonomous agent in Copilot Studio - from defining goals and instructions to connecting business data, automating actions, and incorporating human oversight at critical decision points.

AI agents are moving fast from demos into production systems - and with that comes a new set of questions every technology leader needs to answer: What can an agent do on its own? What requires human approval? And how do you prove what it did?

In this session, we will explore how to design and deploy AI agents that do not just answer questions, but take real operational actions - while keeping humans in control of every critical decision. Expect a live demo on Azure AI Foundry that will show exactly what this looks like in practice.

Building AI products in healthtech requires more than powerful models. This session explores how to turn emerging AI capabilities into useful, trustworthy healthcare solutions, balancing clinical needs, data privacy, regulatory demands, and user experience. Discover practical approaches to moving from promising prototypes to products that deliver meaningful value for patients.

This presentation explores the transformative role of NVIDIA models, advanced techniques such as multi-token prediction and sovereign AI Factory infrastructure in the modern Enterprise sector. It will demonstrate how organizations can leverage high-performance AI technologies to accelerate AI workloads while maintaining full control over their data and ensuring compliance with organizational and regulatory requirements.
In addition, attendees will gain insights into the practical applications and business value of these technologies across various enterprise scenarios. As a special benefit, all conference participants will be offered free access to test and use one of the most advanced AI models available today, providing a unique opportunity to experience its capabilities firsthand.

The success of an AI agent depends not only on its design and tools, but also on the Large Language Model (LLM) that powers it. With an increasing number of models available - from GPT and Claude to Gemini and open-source alternatives - selecting the right LLM requires balancing factors such as reasoning ability, cost, speed, context window, security, and enterprise readiness. In this session, we will explore a practical framework for evaluating LLMs and learn how to match the right model to the right agent use case.