IBM Ventures backs Komodor's push to run AI agents in production
What's the deal? IBM VenturesDealroom has a profile for this one. Try Dealroom → has invested in Komodor, the maker of what it calls an Agentic Operations Platform for Production. The investment, announced on November 10, 2026, backs Komodor's shift from Kubernetes troubleshooting to deploying and governing AI agents across SRE, platform engineering, and DevOps workflows. IBM did not disclose the amount.
Why now? As enterprises operationalize AI, the bottleneck has moved from development to production, where applications and AI workloads run at scale. IBM says infrastructure complexity is growing faster than most platform teams can manage, leaving a gap in reliability, performance, and cost control.
What's the endgame? Komodor launched its Agentic Operations Platform in September, built on six years of solving production problems in mission-critical environments. It lets teams build, run, govern, and optimize AI agents in production, moving operations from reactive troubleshooting to autonomous execution.
What could go wrong? The challenge is scale: deploying a single agent is straightforward, but running many reliably across production systems is harder. Gartner expects more than 40% of agentic AI projects to be cancelled by the end of 2027, citing escalating costs, unclear business value, or inadequate risk controls.
The signal: IBM's move lands in a crowded field — PagerDuty introduced its Paige SRE agent in November, and New Relic announced a Kubernetes troubleshooting agent, both targeting faster incident response. The bet reflects a wider view that operational reliability and human oversight, not just model development, are becoming core enterprise-AI needs.
Read more: ibm.com
Image credit: Generated with Gemini
