Back to Headlines
Tech
May 28, 2026
Analyzed by Glm 4.7 Flash

The Shift from Excitement to Stability: Why Enterprise AI Deals Stall

AI Summary
At TechCrunch Disrupt 2026, Databricks' Arsalan Tavakoli-Shiraji argues that enterprise AI deals are failing not due to poor model performance, but because startups prioritize excitement over operational stability. The market is shifting from experimentation to a phase where governance, integration, and risk management determine success.

The Shift in Enterprise AI Evaluation

Enterprise organizations are not rejecting AI; they are rejecting operational instability. This marks a critical inflection point where the market moves from experimentation to deployment safety. The defining reality separating scalable AI companies from stalled ones is no longer the power of the demo, but the safety of the deployment.

TechCrunch Disrupt 2026: The Operational Pivot

At TechCrunch Disrupt 2026, taking place October 13–15 at Moscone West in San Francisco, Arsalan Tavakoli-Shiraji, co-founder and SVP of field engineering at Databricks, will address this shift during his session on the AI Stage.

  • Event: TechCrunch Disrupt 2026
  • Date: October 13–15, 2026
  • Location: Moscone West, San Francisco
  • Session: "The Enterprise Isn't Broken. Your Assumptions About It Are."

The Pilot Trap: Why Models Don't Fail, Deployments Do

The enterprise AI market is currently plagued by successful pilots that never become real deployments. This failure is rarely due to technical underperformance but rather the organization's inability to absorb the operational consequences of adoption.

Enterprise buyers are no longer asking if AI is exciting; they are asking if it is safe to deploy broadly. Key concerns now include workflow friction, governance, and explainability. An AI product can perform exceptionally well in a controlled environment but fail commercially if it creates instability within the business.

Uncertainty Reduction as a Competitive Advantage

Startups gaining traction inside large organizations share a common trait: they reduce uncertainty. While breakthrough demos generate initial excitement, operational adoption requires clean integration, ease of governance, and trust over time.

Tavakoli-Shiraji brings a unique perspective to this challenge, bridging the gap between enterprise strategy (via his background at McKinsey & Company) and deep technical systems architecture (PhD in computer science from UC Berkeley).

The Future of Enterprise AI Adoption

The next phase of enterprise AI success will likely belong to companies that understand how organizations absorb change, rather than those with the most advanced models. Founders must prioritize long-term operational adoption over initial excitement.