Kosmos Launches Slack Integration to Bring Risk Detection and Root Cause Analysis Where People Work

New integration lets teams review evidence, confirm or dismiss risk events, and track root cause analysis directly in Slack

CHICAGO, Sept. 14, 2026 /PRNewswire/ -- Kosmos, The AI Escalation Platform focused on ending escalation chaos, today announced a new Slack integration that brings its risk event workflow directly into the channels where support and engineering teams already work. When Kosmos detects a new risk event, it posts a summary of what happened and the supporting evidence to a designated Slack channel, where teams can confirm or dismiss the event without switching tools. The new integration is being unveiled in conjunction with Dreamforce 2026.

When a confirmed risk event is promoted to root cause analysis (RCA), Kosmos also updates the team in Slack and delivers the RCA summary there. Every step follows the trust model at the core of the platform: machines propose, humans confirm. The result is a workflow that takes teams from detection through human validation and root cause without requiring them to constantly monitor another application.

"Teams shouldn't have to live in another dashboard just to understand what is happening across their systems," said Sanjay Gidwani, CEO and founder of Kosmos. "When Kosmos sees something important, it should find you, bring the evidence with it and let you make the decision right where the conversation is already happening. Slack is where many of those conversations happen today."

Meet Teams Where They Work
Kosmos connects signals across the systems enterprises already run: customer cases in Salesforce, code changes in Jira and GitHub, and service incidents from observability platforms including Datadog, Grafana and Splunk. The platform correlates those signals to surface potential risk events and structured root cause, with every correlation proposed by machine and confirmed by a human before it becomes part of the permanent operational record.

The Slack integration extends that human-in-the-loop model to where teams already work. Rather than requiring a support or engineering lead to open Kosmos to review a proposed event, the relevant evidence arrives directly in a designated Slack channel. The team can confirm or dismiss it there, and the decision is automatically synchronized with Kosmos.

If the event warrants deeper investigation, teams can promote it to RCA and continue to follow the investigation from Slack. Once Kosmos identifies the root cause, the RCA summary appears in the channel as well, keeping the people responding to the incident informed from initial detection through resolution.

Operational investigations rarely happen in one system. Support may see a customer symptom in Salesforce, while engineering sees a related code change in Jira or GitHub and an operations team sees an infrastructure event in an observability platform. Someone has to determine whether those events are connected, and that knowledge often disappears once the incident is resolved. Kosmos creates a human-confirmed record of those relationships so teams do not have to reconstruct the same incident from scratch when it happens again.

"Machines propose. Humans confirm. Agents act," Gidwani added. "That order is the whole point. We want to make the human confirmation step as natural as possible, and that means putting it into the tools people are already using to work together."

Expanding Visibility Inside Salesforce
Kosmos is also introducing a general preview of Salesforce Error Monitoring, extending its visibility into failures happening inside Salesforce itself. The new capability translates Salesforce internal execution signals, including Flow failures and governor limit errors, into the same observability language as the rest of the enterprise stack. Instead of remaining isolated inside Salesforce until a customer or employee notices a downstream problem, those failures can be surfaced alongside related customer cases, code changes and infrastructure incidents.

"When you run critical workflows on Salesforce, solving a problem once isn't enough if the context behind that resolution disappears with the ticket," said Trey Roldan, CEO at Entrada. "We've seen how much time teams can spend reconstructing what happened across systems, especially when an issue resurfaces months later. What's compelling about Kosmos is that it preserves that operational history and now, by capturing failures happening inside Salesforce itself, it can give teams both earlier visibility into a problem and the context they need to understand it faster."

Salesforce, Dreamforce, Slack and others are among the trademarks of Salesforce, Inc.

About Kosmos
Kosmos is the AI Escalation Platform for enterprise support and engineering teams. Its AI-native platform connects customer cases, code changes, and service incidents across Salesforce, Jira, GitHub, ServiceNow, and leading observability platforms into a single, human-confirmed record, so every agent and every team an enterprise deploys works from the same trustworthy history instead of starting over. Headquartered in Chicago, Kosmos was founded by CEO Sanjay Gidwani, who brings two decades of experience in enterprise delivery operations and the Salesforce ecosystem. Learn more at kosmoslabs.ai.

SOURCE Kosmos AI Labs Inc