The workshop covers building AI agents using the Microsoft stack, focusing on enterprise requirements like security, deployment, and scale.
The hype around AI has created unrealistic business expectations that clash with the non-deterministic nature of current large language models.
Enterprise AI projects face significant delays from compliance and security reviews like GDPR, often due to impractical or outdated rules.
Blocking modern AI tools is counterproductive and leads to security risks, so enterprises should instead enable them securely through proper processes.
The workshop focuses on building real solutions and understanding current AI limitations, such as orchestrating multi-step processes with state persistence.
General-purpose models fail at business process automation because they lack specific domain knowledge, requiring agents connected to internal data.
Attending large developer conferences provides a unique opportunity for networking, learning from diverse perspectives, and experiencing the energy of a live event.
Jobs that call for the skills explored in this talk.