Shy Ruparel

Developer Advocate at Temporal

agentsenterprisedurable-executiontemporal

About Shy Ruparel

Shy Ruparel is a developer advocate at Temporal, the company behind the open-source durable execution platform of the same name. He speaks to engineers building AI agents about running them reliably in production.

At the AI Engineer conference in October 2026, he presented Temporal's second annual State of Development survey of 554 engineers and engineering leaders who actively use AI agents, and demoed an agent-assisted payout workflow that keeps its state and its AI work through repeated worker crashes.

Career Highlights

  • Developer advocate at Temporal
  • Presented Temporal's 2026 State of Development survey on AI agents in production at AI Engineer

Notable Positions

On the Four Gaps

Ruparel frames the survey as four gaps: daily agent use (80.8%) versus production use (49.1%), trust in agent output (85.5%) versus daily failures (41.1%), fast prototypes versus months-long rollouts at companies with 5,000+ employees, and seeing failures versus recovering from them.

On Planning for Failure

Teams that call themselves successful are 1.7 times more likely to hit agent issues continuously, so he argues the goal is not zero failures but failures that cost nothing: completed steps and model calls should survive a crash instead of being rerun.

On Bounding the Agent

In his payout demo the deterministic workflow owns the money and the state, and the agent only advises. If the model fails, the payout continues without the agent's brief, and a human approves the payment.

Key Quotes

  • "failure isn't going away. You have to account for it. You have to plan for it." (on agent reliability)
  • "if the model had failed, the payout would have continued without the brief." (on keeping agents bounded)