Seekr

Seekr is a leader in explainable and trustworthy AI for high‑stakes operational decisions across global enterprises, government, and regulated industries. We provide secure, auditable AI products and solutions for organizations that need transparency, accuracy, and compliance at the core of how they deploy AI in the real world. SeekrFlow, our AI Operating System, enables leaders to build, govern, and deploy domain‑specific large language models and AI agents with their own data while reducing bias and simplifying oversight. As a SOC 2 compliant company, Seekr delivers AI that meets the highest standards of security, regulatory compliance, and operational control. Learn more at www.seekr.com.

View websiteBooth-2005

Previous Sessions

HumanX 2026 San Francisco

Beyond the Black Box: Building AI Systems Humans Can Actually Trust

As AI systems evolve from narrow task-specfic models to autonomous, agentic architectures capable of perception, reasoning, and execution, trust becomes an operational requirement. The question facing leaders is no longer whether AI can perform, but whether its decisions are inspectable, governable, and aligned under real-world conditions. In this session, Stefanos Poulis, Chief Technology and AI Officer at Seekr, draws on Seekr’s work in explainable AI (xAI), decision transparency, and human-in-the-loop system design to outline a modern blueprint for trusted agentic systems. He will unpacks why static governance frameworks and post-hoc oversight mechanisms break down in adaptive, multi-agent environments and how embedding contestability, steerability, and structured human feedback directly into system architecture enables continuous oversight, calibrated autonomy, and durable alignment. This talk bridges technical rigor with operational reality, offering a systems-level perspective on building AI that organizations can confidently deploy, audit, and scale. From this session, you will learn how to: Diagnose where traditional AI governance models fail when autonomous agents make and execute decisions end-to-end. Design agentic systems with built-in contestability and steerability, enabling humans to challenge, override, and guide model behavior over time. Operationalize transparency and feedback loops within production AI workflows to support monitoring, auditing, performance assurance, and continuous improvement.

Best for:

Technology