RemediAI · per-agent add-on
An ops assistant that stays inside your estate—and your safeguards.
RemediAI is Remedion's licensed troubleshooting chat. It searches assets you can see, explains failed jobs, and drafts scripts with the same approval and canary rules as the rest of the platform. Hard denies are enforced in code, not only in the prompt.
Event-log action plans and the fleet script generator remain on AI operations in Base. RemediAI is the conversational operator that sits on top—licensed per agent.
RemediAI
Scoped to client Quadris · 3 assets in view
Operator
Why did the Dell firmware job fail on SRV-DC-01, and what should we try next?
RemediAI
Job 1842 stopped at verify: the package downloaded, but the reboot-pending flag stayed set. Last event log is Windows Update 19 / 0x80240034. Draft a canary script to collect CBS.log excerpts—do not deploy estate-wide until you approve a single-host pilot.
Use cases
Built for the questions operators actually ask.
Why a job failed
Ask why a software, firmware, or remediation job stalled on an in-scope asset. RemediAI grounds answers in job records and recent event logs—not generic advice.
Log and incident triage
Pull the latest Windows events for a host, summarise what matters, and propose the next investigation step without leaving the portal.
Draft fix scripts
Describe the outcome in plain English. RemediAI can propose a PowerShell or bash draft. Mass run is blocked: a single-host canary and human approval stay mandatory.
Product guidance
Ask how to use Remedion itself—policies, application control, iDRAC, Gateway, or reports—scoped to features your licence actually includes.
Offline and gateway-site helpers
Helpers still work when sites have no inbound ports. Agents and gateways poll outbound HTTPS only; RemediAI drafts jobs the relay can pick up without opening the firewall.
Safeguards you can show an auditor
RemediAI cannot skip canary, approve its own scripts, or act as a general-purpose chatbot against your infrastructure. Refusals are logged; answers stay inside client scope and licensed modules.
- No MSI builds, agent packaging, or source/infra changes
- No database access, dumps, or destructive SQL
- No privilege escalation or role changes
- No secrets, enrolment tokens, credential vault, or web shell
- No cross-tenant or out-of-scope assets
- Scripts stay drafts until an operator canaries and approves them
Works with Gateway sites
Locked-down offices do not need inbound ports for AI help. Operators chat in the portal; proposed jobs wait for the agent or gateway to poll. The same hard denies apply on air-gapped-feeling networks as on the open internet.
See RemediAI on your estate
Walk through scoped chat, job analysis, and draft-only scripts with the same approval rules as production.