Droma, the leader in predictive and agentic AI for IT operations, today announced the general availability of the Droma Problem Manager Agent and the Droma SRE (Site Reliability Engineering) Agent. The two role-based agents work alongside Droma's previously launched autonomous L1/L2 Operator Agent to form a closed loop across IT operations, resolving incidents as they happen, eliminating the causes of recurring failures, and preventing new failures by design.
The launch is Droma's first since the company rebranded from Grokstream, with its Grok® platform now the Droma platform. Droma customers—across IT and network operations teams, in industries such as telecommunications, financial services, energy, data centers, and managed service providers—keep the same team, platform and support under the new name.
Designed for Delegation Via Trusted Agentic Operations
All Droma agents run on the Droma Cognitive AI Engine, which combines memory, reasoning, prediction and continuous learning into one shared operational intelligence. Rather than one general-purpose AI assistant, Droma gives each operational role its own agent, and the shared intelligence lets them work as one team.
Every incident the L1/L2 Operator Agent resolves becomes evidence for the Problem Manager Agent, every recurring cause it uncovers becomes a design signal for the SRE Agent, and every fix the SRE Agent drives means fewer incidents reach the front line. Each turn of the loop compounds: incident volume falls, resolution gets faster, and engineering time shifts from firefighting to improvement. The loop builds on a platform that already compresses operational noise 2 to 3 times beyond rule-based approaches.
“Trust makes that delegation possible,” said Casey Kindiger, CEO of Droma. “Droma agents explain reasoning, show evidence, and know when to involve a human. Organizations then decide how much to hand over by role, action and risk, moving from AI assistance to human-approved actions to autonomous operations as confidence grows.”
The Droma Problem Manager Agent
While most AI solutions for IT operations stop at the incident, problem management—the ITIL (Information Technology Infrastructure Library) practice that prevents the same outage from happening again—is where recurring incidents are eliminated. Yet in many enterprises problem management is understaffed or left undone.
The Droma Problem Manager Agent aligns with ITIL practices, analyzing the full pattern of recurring failures and drawing upon related incidents, historical changes, and past resolutions—not just individual alerts.
The agent works each recurring pattern through five stages already familiar to problem managers: problem overview, root cause analysis, impact assessment, workaround, and assignment and ownership. It explains what recurs and how often, weighs potential root causes with the evidence for each, judges impact, separates today's workaround from the permanent fix, and identifies the owning team, packaged as a complete handoff ready for the problem record.
The Droma SRE Agent
The SRE Agent continuously evaluates architecture, capacity and dependencies to find the conditions that let failures happen, models trends against SLOs (service level objectives) and error budgets to flag services heading toward a breach and ranks resilience work by potential blast radius. By moving reliability engineering from firefighting to foresight, the result is design-level reliability thinking across every service, without adding SRE headcount.