Bito CEO Amar Goel sits down with Kai Waehner, Global Field CTO at Confluent, to explore the critical role of real-time data in the GenAI revolution.
Kai explains why so many AI projects stall after model training, how modern data architecture unlocks real business value, and why operational data is the missing piece in reducing hallucinations.
With years of experience helping enterprises design streaming-first systems, he shares a clear-eyed look at where AI breaks down today and what patterns will shape the next wave.
Episode 11 highlights
In this episode of Peer Review, Amar and Kai discuss:
- The deployment dilemma: why so many AI models never make it into production.
- Data vs models: why fresh, contextual data from operational systems is more valuable than chasing the next LLM.
- Predictive vs generative AI: how real-time predictive patterns like fraud detection are being adapted for GenAI.
- The “Shift Left” architecture: Kai’s perspective on ensuring high-quality data across all systems and avoiding the reverse ETL trap.
- The future is Agentic: why agentic AI is the next wave and how data streaming forms its natural foundation.
Watch PEER REVIEW
Don’t miss out on future episodes. Subscribe to Bito’s YouTube channel, follow us on LinkedIn, and join us on Reddit to stay updated.
by the Bito team. 
