The AI in Business Podcast

The consistency gap in enterprise AI represents a critical failure point where unpredictable system behavior outside of controlled demos threatens to derail executive sponsorship and regulatory compliance. In this episode, Amar Akshat, SVP & Chief Architect at Paysafe, examines how leaders can move beyond experimental shadow AI by embedding determinism and high-threshold guardrails directly into the production pipeline. The discussion outlines a rigorous evaluation framework centered on treating prompts as versioned intellectual property, implementing Know Your Agent (KYA) policy envelopes, and ensuring every agentic decision remains holistically auditable. Learn how brands work with Emerj and other Emerj Media options at https://go.emerj.com/partner

Legacy financial systems often trap organizations in "data swamps" where AI is mistakenly treated as a magic fix for fundamentally broken manual architectures. In this episode, Juan Orlandini, CTO of North America at Insight, outlines why senior executives must distinguish between statistical AI outputs and the mathematical precision required for financial compliance to avoid significant reporting risks. The conversation provides a roadmap for building a scalable operating layer by prioritizing data engineering and leveraging established vendor knowledge to protect long-term investment. This episode is sponsored by K1x. Learn how brands work with Emerj and other Emerj Media options at https://go.emerj.com/partner

R&D teams are starting to advance AI capabilities faster than they can translate them into measurable business value, creating mounting friction between scientific progress and operational reality. In this episode, Aziz Nazha, Global Head of AI Innovations Institute at Incyte Pharmaceuticals, examines how culture, talent, infrastructure, and expectation‑setting determine whether AI meaningfully improves drug discovery and development. He highlights the practical shifts required — from redesigning workflows to disciplined upskilling and targeted validation cycles — to ensure AI adoption accelerates cycle times rather than getting stalled by organizational bottlenecks. This episode is sponsored by Deloitte. Learn how brands like Deloitte work with Emerj and other Emerj Media options at go.emerj.com/partner

A recurring challenge for leaders is that the use cases they expect to automate rarely match what customers actually struggle with once large‑scale conversation data is analyzed. In this episode, Shezan Kazi, Head of AI Transformation and AI Products at Dialpad, examines how autonomous agents should take the first pass on high‑volume deterministic requests, when they must hand off to humans, and why confidence scoring and oversight models are essential for safe deployment. He highlights the practical steps leaders can take — from starting with low‑risk, high‑impact tasks to redesigning processes around real interaction data — to expand automation responsibly and improve customer outcomes over time. This episode is sponsored by Dialpad. Learn how brands work with Emerj and other Emerj Media options at go.emerj.com/partner

A major shift is underway as enterprises move from lab‑ready computer vision to the far more complex reality of deploying visual intelligence across messy, variable, high‑stakes physical environments. In this episode, Joseph Nelson, Co‑founder and CEO at Roboflow, examines how dependable visual data, models tuned to real operating conditions, and integration with existing production and safety systems determine whether visual AI delivers meaningful value. He highlights the practical moves that matter most: securing consistent visibility into key processes, choosing a first deployment that proves impact, and scaling only once the operational foundations are in place. This episode is sponsored by Roboflow. Learn how brands work with Emerj and other Emerj Media options at go.emerj.com/partner