Data manager · DataHub developer
Shipped (V0)Business rule validation & testing
On every rule save, DataHub chains governed agents: validate syntax and best practice, evaluate against business assertions, build and run tests.
The full BR quality pipeline — validator, evaluator, tester — is implemented in the platform catalog and runs in task mode, on contract-tested, fixture-backed DataHub tools until live connectors land in Beta.
In DataHub
The daily reality
The work this replaces
A rule that parses is not a rule that is right. Reviewing syntax, best practice and actual behavior against assertions is manual, inconsistent and skipped under deadline — until a bad rule reaches the golden copy.
How it works
What the agent actually does
dh-br-validator · dh-br-evaluator · dh-br-tester
Triggered by the workflow
On every 'BR saved' event, the DataHub workflow chains task agents. The application orchestrates; each agent is a governed step.
Validate
dh-br-validator checks syntax and best-practice compliance and returns a prioritized improvement plan. The pipeline stops on high-severity issues.
Evaluate
dh-br-evaluator runs the rule against business assertions on the DataHub engine: pass or fail per assertion, a coverage score, diagnostics.
Test
dh-br-tester builds and runs a BRTU test suite — generates the cases, schedules execution, reports the results.
Structured verdicts
Every step returns a typed, schema-validated verdict the workflow acts on automatically — no prose for a human to re-parse.
The outcome
Outcome
Every saved rule gets the same rigorous review, with structured verdicts your data workflow acts on automatically.
Governed by design
Agents never orchestrate the pipeline — DataHub does. Each step is a bounded, traced run with its own audit record, and each agent carries a versioned eval baseline.
FAQ
Questions teams ask about this
Who orchestrates the pipeline?
The application. NeoXam Agents deliberately avoids agent-to-agent graph orchestration: the DataHub workflow decides what runs and when; each agent executes one bounded step and returns a typed verdict.
What happens when validation fails?
The validator returns a prioritized improvement plan, and the workflow stops the chain on high-severity issues — a structured outcome the product can act on, not an exception buried in logs.
Is the conversational BR assistant part of this?
It complements it. dh-br-assistant (Beta, V1) lets developers draft, explain, validate and test rules interactively in the authoring screen; actions that touch the live engine require user confirmation.
Go further
Related use cases and concepts
Business rule generation from plain language
Shipped (V0)Write 'flag NAV deviations above 2%' — the generator returns a syntactically valid business rule as a typed object DataHub consumes directly.
Data-quality remediation
In product · unified at GAData-quality errors investigated in the context of the feeds, rules and configuration involved — so remediation becomes review-and-apply.
From the glossary
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.