Welcome
User Guide
Appendix
Best Practices for MaxAI Queries
Best Practices for Segmentation Queries
Generates Athena SQL + JSON criteria for Unica CDP segments. The SQL returns customer_id and JSON defined filters.

HCL MaxAI Documentation
Release Notes
Deployment Guide
Administrator Guide
User Guide
- An Introduction to MaxAI
  MaxAI is an AI-driven offering introduced in 25.1.0 version of HCL Unica+. It is a part of the HCL Unica+ marketing platform.
- MaxAI Assistant
  MaxAI Assistant helps in real-time decisions and optimizations.
- MaxAI Content Generation
  MaxAI Content Generation enables a Marketer to effortlessly optimize the content of their emails in Unica Deliver and Rich Text Fields during Offer creation or Offer modification in Centralized Offer Management. With intelligent AI-driven suggestions, marketers can instantly refine their messaging for better clarity, tone, and impact ensuring that the core message resonates with their audience.
- MaxAI Email Scoring and Insights
  MaxAI Email Scoring and Insights enables a Marketer to improve their email content by scoring it based on various marketing-related parameters. With intelligent AI-driven suggestions, marketers can refine their email communication to make it a professional and high-impact marketing communication. This feature is applicable only for Unica Deliver.
- MaxAI Journey Insights
  Access actionable insights for Unica Journey. In just a few clicks, marketers can move from observation to action.
- MaxAI Reporting Insights
  MaxAI Reporting Insights provides AI-driven insights for contextual and actionable-analysis based on the selected chart or report. With the analysis, you can continue to the chat with the AI in MaxAI assistant, ask follow-up questions to receive appropriate responses enabling seamless exploration of trends, detection of anomalies, and data-driven recommendations.
- MaxAI Session Insights
  MaxAI also integrates with HCL Discover. MaxAI Session Insights provides AI-driven contextual insights and actionable-analysis based on the selected session.
- Appendix
  - Best Practices for MaxAI Queries
    - Best Practices for Analytics Queries
      Transform everyday questions into precise MariaDB SQL queries for Campaign, Deliver, or Journey data.
    - Best Practices for Document Search Queries
      Rephrase your question, fetch docs via embeddings, and get a concise answer.
    - Best Practices for Plan Execution-related Queries
      MaxAI can execute tasks in Plan like List Approvals, Create Approvals, Assign Approvals, Update Approval Status and Modify Due Date, and Assigning Offers or Offer Lists to Approvals.
    - Best Practices for Segmentation Queries
      Generates Athena SQL + JSON criteria for Unica CDP segments. The SQL returns customer_id and JSON defined filters.
    - Best Practices for Journey Queries
      Generate SQL for Journey tables: Performance, Milestones, Goals, and Touchpoints.
    - Best Practices for Subject Line Analysis Queries
  - Queries - Examples
    The following sections contain examples for proper queries that you can ask MaxAI related to Unica or HCL Discover.
  - FAQs
    The following are some frequently asked questions related to MaxAI Assistant:
  - System Safeguards
  - How the System Responds to Errors
    This topic explains how the system handles scenarios where it cannot generate a response due to incomplete or unclear inputs. It provides users with insights into fallback mechanisms, error messaging, and clarification requests.

Best Practices for Segmentation Queries

Generates Athena SQL + JSON criteria for Unica CDP segments. The SQL returns customer_id and JSON defined filters.

How to Use Prompts?

Demographics: Age, Gender, Location. Behavior: Transactions, Responses. Timeframe and Thresholds.

Example: Segment female customers aged 25–35 in Germany with >3 transactions and revenue >1000 in last 90 days.

Examples for Good Prompt vs Bad Prompt


Prompt	Good / Bad?	Reason?
Segment male customers age 40–50 in US with no purchase >180 days.	Good	Full criteria.
Segment users.	Bad	Needs details.
Create segment high spenders.	Bad	Missing thresholds.
Segment female, age 30–40, France, >2 orders last month.	Good	Specific date and metrics.
Segment people by event.	Bad	Which event?
Segment single customers in UK with >$500 revenue.	Good	Demographic + revenue.
Make group.	Bad	Lacks context.
Segment Gen Z in CA with clicks >5.	Good	Age cohort + state + behaviour.
Segment behavior.	Bad	Needs specifics.
Create fancy segment.	Bad	Non-descriptive.
Segment customers who dropped off car loan journey	Good	Journey stage + context
Segment users with email opens >2 and click-throughs >1	Good	Specific behavioral metric
Segment by app version	Bad	Missing filter logic (which version?)