Qualys Builds Confidence in AI-Powered Remediation
18m
Confidence Becomes the Remediation Bottleneck
Finding vulnerabilities is only part of the security challenge. Fixing them without disrupting the business is the harder task. At Qualys ROCon 2026, Kunal Modasiya explores why AI-powered remediation needs more than automation. It also needs evidence that each action will reduce risk without creating operational problems.
Modasiya, senior vice president of product management at Qualys, joins Alan Shimel to discuss TruRisk Eliminate 2.0 with TruConfidence. Their conversation centers on the gap between identifying exposures and taking action. As AI accelerates exploit development, they argue that traditional remediation processes cannot keep pace.
Remediation Extends Beyond Patching
Patching is not always the best response to an exposure. Modasiya outlines alternatives such as isolating applications, applying temporary mitigations, and removing unused software. These choices can reduce risk while teams evaluate a permanent fix. They can also avoid unnecessary disruption to essential services.
The discussion highlights network devices that cannot easily accept endpoint agents or tolerate reboots. Modasiya describes custom mitigations designed to address these environments without requiring a patch. He also explains why blocking one vulnerable application may be preferable to isolating an entire machine.
Unused software offers another practical opportunity. Removing an outdated application can eliminate an exposure rather than preserve it through repeated patching. The broader goal is choosing the right response for each asset and its business context.
Agentic Execution Meets Operational Checks
Modasiya describes Agent Sara as a digital worker that helps plan and execute remediation strategies. The agent gathers intelligence, recommends actions, and seeks permission before orchestrating changes. AI-powered remediation intelligence adds patch reliability scoring and failure analysis to support those decisions.
The interview also examines a Dynatrace integration for checking system health before and after patch deployment. Modasiya explains a staged rollout approach that starts in a lab before expanding into production. Health indicators help teams identify instability and determine when a rollback is needed.
For security and IT leaders, the takeaway is practical: speed alone does not establish trust. Remediation must account for business continuity, application dependencies, and measurable results. The conversation connects faster execution with the operational confidence needed to put it into practice.