What is NeoXam Agents?
An overview of the governed AI agent layer embedded across the NeoXam product suite.
NeoXam Agents is a unified AI agent layer for investment management operations, embedded across the NeoXam product suite — ARO, DataHub, GP, Impress, and PMS. Instead of a separate chatbot beside your tools, agents are bounded, declared capabilities that your products invoke exactly where reasoning is needed: investigation, interpretation, multi-step context handling, and conversational assistance.
What agents do — and what they don't
Agents handle bounded reasoning tasks: investigating a reconciliation break, drafting a business rule, explaining an exception. The application decides what runs and when; the agent decides how, within a bounded task. Agents never orchestrate business workflows and never write to your books — deterministic systems of record stay in charge, and your teams stay in command.
Two ways to run
Task mode is a stateless call from a business workflow: typed input in, schema-validated output back, with hard ceilings on duration and tool calls. All agents shipped today run in task mode.
Conversational mode is a multi-turn chat session embedded in a product screen, with platform-managed memory and streamed answers.
Beta
Conversational agents are rolling out with the V1 Beta. The first pilot is the DataHub assistant.
Where the platform stands today
The V0 platform is delivered: the ARO reconciliation investigator runs end-to-end on the governed runtime, alongside the DataHub Business Rule agent family. Every agent is declared in a Git-backed Catalog, gated by evaluation baselines, policy-controlled at every tool call, and fully traced. General availability is planned for Q3 2026, and the Early Adopter Program is open now.
Still stuck?
Tell us what you were trying to do and where it failed — screenshots and run IDs help. A human follows up.