This factsheet is available in English. Catalogue language:
Classified under the CAI Score™ · L3 · Operate

Reconciliation Root-Cause Agent

Investigates each reconciliation break, prioritises it by materiality, and narrates why it happened and how to fix it.

CAI tier L3 Operations & Service Delivery Operations · Reconciliations

Where it fits

Sits at the front of the reconciliation queue: it triages breaks the moment they appear, before an analyst opens the first one, and feeds clean, explained breaks into the dual-control step.

Value & ROI

Cuts manual investigation effort and stops systemic issues from recurring by spotting the pattern, not just the symptom.

KPI it moves

Lower break-resolution time · fewer repeat breaks · cleaner audit trail.

Risk & compliance posture

L3, operates inside the recon workflow; routine breaks are explained automatically, anything material is routed to a human. Every investigation produces audit-ready evidence.

Pain points it removes

What the L3 · Operate tier means

Executes routine, low-risk actions inside a bounded workflow: classify, route, fulfil, log. Exceptions and anything unusual are handed to a human. Every action is time-stamped. Human role: Owns exceptions.

How you adopt it

Ships pre-built for Microsoft Copilot Studio, Power Automate and Azure. You connect it to your environment, classify it on the CAI Score, and it runs in your own tenant. Customer business content remains in the customer-controlled Microsoft environment; ColleagueAI processes separate governance-layer metadata as described in the Trust Center.

Deployment & data

Agents run inside your own Microsoft Copilot Studio, Power Automate and Azure environment. Colleague AI hosts only the control plane: scores, policies and audit metadata. Customer business content remains in the customer-controlled Microsoft environment. ColleagueAI processes separate governance-layer metadata as described in the Trust Center.

What you get

One-time purchase. Access to your connect packages is arranged after your access request is approved and commercial setup is confirmed.

Price

$9,900 one-time for a perpetual licence to this version, plus $149 / month for Continuous Certification — recommended, and the only route to updated certified releases.

The version you buy is yours forever. The rules it was built against are not: without an active subscription a newer certified release is a new licence purchase, and you follow the covered scope yourself. With it, the package stays listed as a current Colleague AI Certified Release, you get the updated release when a change inside the covered scope requires one at no further licence fee, and you get a Certificate ID a reviewer can check.

Colleague AI’s own programme — not third-party accreditation, and it does not certify how you use the agent. What it covers.

About Colleague AI

Colleague AI is the trust layer for enterprise AI. It classifies AI agents against the CAI Score (a five-tier risk classification from L1 to L5) documenting each agent’s controls and producing an audit trail. Agents run inside your own environment; we host only the governance control plane. So you deploy AI you can defend.

Request access → Request a demo ⤓ Agent dossier (PDF)