HCL MaxAI CDM offers a a unified, extensible, and vendor-agnostic data layer designed to support MaxAI across the entire Martech landscape.

The new features in the first release of HCL MaxAI CDM are as follows:

If an issue cannot be resolved by documentation, your designated support contact can log a call with HCL Technical Support.

This chapter explains how data is organized within the CDM and how it flows through different layers. It helps users understand where data comes from, how it is standardized, and which layer should be used for consumption.

HCL MaxAI CDM integrates with the following Unica applications:

This section defines the operational scope, responsibilities, and role boundaries for teams running day-to-day Canonical/UDS pipeline operations.

This section is for operational teams responsible for executing documented runbooks for recurring tasks and issue resolution.

This section covers daily monitoring responsibilities, log access, and first-response actions for typical pipeline failure modes.

This section provides backend-specific operational awareness for multi-database deployments. DBA teams may also reference sizing tables herein.

This section helps operations teams recognize performance deviations and know when to escalate. DBA teams own tuning decisions.

This Adoption Guide explains, in plain language, how to perform that mapping exercise end-to-end, from understanding your source data all the way to validated, signed-off mapping rows ready for metadata loading and ETL generation.

A CDM onboarding project can only succeed if the mapping exercise is done well. Poor mapping has cascading effects across data quality, ETL generation, analytics, compliance, timeline, and MaxAI readiness.

CDM onboarding follows a structured lifecycle. The interface mapping exercise sits at the heart of the Discovery and Design phase — after source understanding and before any ETL or DDL is generated.

The mapping exercise requires involvement from five key roles, each bringing specific expertise and perspective to ensure the exercise is complete and correct.

Do not open the MID until the following artefacts have been collected from the client. Attempting to map without these leads to guesswork and rework.

You do not need to know the full CDM model in detail before starting. Use the subject area guide to identify which family of LDZ tables your source data likely belongs to.

These principles must guide every mapping decision. Violating them leads to data quality problems, ETL complexity, and rework.

Follow all eight steps in sequence for each source system in scope. Do not skip steps to save time — each step produces decisions that the next step depends on.

Use this decision guide when you are unsure which CDM entity to target. Work through the questions from top to bottom and stop at the first match.

The following patterns appear in almost every CDM onboarding engagement. Recognising the pattern early saves significant time.

The following worked examples use a fictional retail bank, "Horizon Bank", which is onboarding its core banking data into CDM.

Every mapping row in the MID should have at least one data quality rule. Every source column containing personally identifiable information must be flagged in the PII column.

Not every source field will have a direct target in the standard CDM model. Follow a hierarchy of options before raising an extension request.

The mapping is not complete until all three sign-off gates have been passed. The order matters — do not request engineering sign-off before business sign-off.

After the first ETL run, use the following structured debugging checklist to validate that data has loaded correctly from source to LDZ. Work through each check in sequence.

This section addresses frequently encountered issues during the mapping exercise and provides remediation approaches.

Use this checklist to confirm that the mapping exercise is complete and the MID is ready for handoff to engineering. Every item must be checked before sign-off.

Three-Layer Implementation with Metadata as Core