Data Manager · DataHub Developer
AI agents for enterprise data management
Business rules drafted, validated, evaluated and tested in minutes — by agents that know your syntax dictionary, your rule corpus and your engine.
Your data platform now speaks your language.
Pains → gains
From investigation toil to structured findings
Deterministic systems keep the books. Agents handle bounded reasoning tasks inside them. Your team stays in command.
The work today
Business-rule authoring and data-quality remediation are slow, expert-bound and backlogged.
The knowledge lives in a few heads and a thousand pages of documentation.
Every configuration question queues for the same few specialists.
Generic copilots don't know your syntax dictionary, your rule corpus or your engine.
With NeoXam Agents
Rules in minutes, not days
Rules are drafted, validated, evaluated and tested against the real syntax dictionary and rule corpus — and returned as typed objects DataHub consumes directly.
A quality pipeline, not a one-off
On every save, the DataHub workflow can chain governed agents — generate, validate, evaluate against business assertions, test. The app orchestrates; agents return structured verdicts.
Answers with deep links
Configuration questions will be answered from documentation and live configuration, with deep links to the exact screen — arriving with the Beta's conversational assistants.
Quality is a gate, not a hope
Every agent ships with a versioned evaluation dataset and a baseline score that can only go up; CI re-runs evals on every change.
The agents
Agents that matter for your team
Every agent is declared in a Git-governed catalog, gated by evaluation baselines, policy-controlled at every tool call, and traced end to end. Statuses below are exact.
BR Generator
Shipped — task modedh-br-generator
Generates a Business Rule expression from a plain-language specification, grounded in the syntax dictionary and an examples corpus.
BR Validator
Shipped — task modedh-br-validator
Validates a rule on syntax and best-practice compliance and returns a prioritized improvement plan as a structured verdict.
BR Evaluator
Shipped — task modedh-br-evaluator
Evaluates a rule against business assertions on the DataHub engine: pass or fail per assertion, a coverage score and diagnostics.
BR Tester
Shipped — task modedh-br-tester
Builds and runs a BRTU test suite: generates test cases, schedules execution and reports results.
BR Assistant
Betadh-br-assistant
An interactive Business Rule expert inside the authoring screen: draft, explain, validate, test and improve rules through conversation. Actions touching the live engine will require explicit user confirmation.
DataHub Assistant
Betadatahub-assistant
A product assistant grounded in documentation and live configuration — "How do I configure a Kafka feed?" — answering with deep links to the exact screens. The Beta milestone pilot.
Key capabilities
Why this works here
Plain language in, valid rule out
ShippedWrite "flag NAV deviations above 2%" and receive a syntactically valid Business Rule drafted from the syntax dictionary and example corpus.
Structured verdicts
ShippedOutputs are schema-validated typed objects. A malformed answer is a hard failure, never a silent success — so the data workflow can act on verdicts automatically.
Grounded in your documentation and configuration
BetaConversational assistants will retrieve from product documentation, configuration objects and the BR example corpus, with deep links to product screens.
NeoXam products
Where these agents live
Agents are embedded in the NeoXam products your teams already use — not bolted on beside them.
- Shipped — task mode
DataHub
Enterprise Data Management
Business rules drafted, validated, evaluated and tested by agents that know your syntax dictionary, your rule corpus and your engine.
FAQ
Questions your team will ask
Which DataHub agents are real today?
Four task-mode agents are shipped in the catalog — generator, validator, evaluator and tester — running against contract-tested fixture tools today, with the live engine connection landing in Beta. The conversational BR Assistant and the DataHub product assistant follow in Beta.
Do the agents learn from our data?
No. There is no fine-tuning on client data. Agents are grounded at run time in your syntax dictionary, rule corpus, documentation and configuration, and every output is traced to the run that produced it.
How is rule quality controlled?
Outputs are schema-validated typed objects — a malformed rule is a hard failure. Each agent carries a versioned eval dataset and a baseline score that can only rise; CI re-runs evals on every change to an agent, skill or bundle.
Can agents change our production rules on their own?
No. Task agents return drafts and verdicts that your workflow and your people act on. Any action touching the live engine will show a confirmation step first, shipping with the Beta's human-in-the-loop approvals.
Join the Early Adopter Program
General availability lands in Q3 2026. The Early Adopter Program is open now — a limited cohort, a one-year platform trial, and three workshop streams: Business ROI, Compliance, and Operational fit.