STAGING This is not the live site
Teammates | centrexIT Knowledge Center
Loading...
centrexIT
Knowledge Center

Teammates

Reference 1 min

The centrexIT Knowledge Center is supported by focused AI agents and integrations that maintain content quality, orchestrate publishing, and connect the KB to operational systems. The launch architecture keeps Knowledge on its own API, database, and search indexes so article lifecycle, recommendations, and agentic retrieval can evolve without depending on the broader centrexAI-core platform.

Aeris is the quality and evaluation agent. She scores articles against a set of quality criteria, identifies exemplary articles that serve as benchmarks, and provides targeted suggestions for content improvement. Aeris helps maintain consistency and standards across the entire Knowledge Center, giving authors clear guidance on how to raise the bar. When you see a quality score on an article, Aeris is the one behind it.

Atlas is the build orchestration and intelligence agent. He manages the content pipeline, handles intent detection, and coordinates between the KB and other systems. Atlas uses the Knowledge API, Azure AI Foundry, and Azure AI Search to power smart search, article recommendations, and RAG-backed article drafting.

The Knowledge API is the integration backbone for article CRUD, review, approval, publish jobs, search traversal, recommendations, templates, feedback, notifications, and audit events. Frontend authoring tools, Atlas, GitHub publishing workflows, and internal automations should call this API directly with Knowledge-scoped credentials or Entra-backed user context.

The Halo integration serves as the ticketing bridge between the Knowledge Center and HaloPSA. It helps turn resolved support issues into candidate articles, then routes those drafts through the Knowledge governance lifecycle before publishing.

Next integration work should focus on production Foundry tool wiring, Azure AI Search relevance tuning, Rewst orchestration, Teams surfacing, and automated article recommendations from recurring ticket patterns.