AI Workflows in CMA

Practical AI in CMA

Help staff turn CMA context into useful next steps.

Ask CMA Assistant to explain your workflows and help prepare the work. Connect optional AI tools to calls, client records, tasks, notes, documents, and reports. Your team reviews the result and decides what happens next.

  • Connected to CMA context
  • Configured for the job
  • Reviewed by a person
Work with context

Use information connected to the client and the job.

Continue the workflow

Move reviewed output into notes, tasks, documents, or reports.

Keep people in control

Require a person to interpret the result and choose the action.

CMA Assistant

Understand how your CMA is set up to work.

We call this Reflective Intelligence: CMA Assistant connects application knowledge with approved settings from your organization to explain how the pieces fit together. Staff gain a starting point for onboarding, refresher training, and discussions with administrators and support.

Trace a configured workflow

Show me the task flow for lead to client enrollment.

Explore related task definitions, outcomes, phase rules, and agency or company defaults. A diagram can clarify supported branches and relationships alongside the explanation.

Understand the paperwork

Which documents belong to our DMP Enrollment package?

Connect package contents with document templates, task links, and client portal steps. Ask another question to explore a condition or compare named configurations.

Check operational settings

What payment minimums and proposal requirements are configured for this creditor?

Review approved creditor settings and related terms. Other supported questions cover payment hold settings, stored bank holidays, and fee schedule definitions.

Keep the conversation connected to the work

Type without enabling a microphone, speak when useful, or choose to hear answers aloud. Ask for more detail, adjust reasoning strength for a complex question, then request a supported note or task draft in the same conversation.

Read about CMA Assistant

Review the explanation before acting

Availability depends on configured AI services, Knowledge Base access, and user permissions. Approved configuration reads respect agency and company scope. Answers identify missing or ambiguous information; staff verify material details. A configured relationship does not establish client eligibility or confirm that a step ran. People save record changes through CMA’s normal controls.

AI across the operation

Use AI to support everyday CMA workflows.

These capabilities cover workflow questions, navigation, live and recorded calls, task events, documents, and reporting.

Authorized CMA staff

CMA Assistant

Ask about the workflow, then prepare the next step.

Type or speak to find permitted pages and clients, understand approved local settings, or ask about limited visible page content. CMA Assistant can also prepare supported List filters, fill supported note or task fields, and create editable drafts.

Staff confirm prepared filters before the list runs and review and save record changes through normal CMA controls.
Counselors and service teams

Call Assist

Keep important points visible while the conversation is happening.

When enabled for an authorized user, Call Assist can show a live transcript for an eligible conversation, plus cues, checklists, summaries, and a note draft based on the selected prompt.

The counselor stays responsible for the conversation, the client record, and every next step.
Service and quality teams

CallLog Transcriptions

Keep a reviewable transcript with the CallLog record.

When recordings and transcription services are configured, CMA can retrieve eligible call recordings and store the transcript on the CallLog. Staff can search and review the text with call details and use an approved prompt to analyze it.

The transcript can support call review, a Client Note, or a configured Smart Notes workflow.
Client service teams

Smart Notes

Create a Client Note from an eligible call transcript.

When configured, Smart Notes applies an approved prompt to an eligible transcript and saves the generated result in the client's Notes area under the selected Note Type. Each note links to the source Call Log for review.

Staff compare material details with the transcript or recording and add clarification when needed.
Workflow administrators

Task AI Analysis

Run a configured analysis when a task is created or completed.

An active AI prompt template linked to the Task can use permitted client and workflow context. CMA sends the generated prompt through its configured AI service and stores the result as an AI ClientMessage for staff review.

Agencies can connect repeatable analysis to a defined point in the task workflow.
Counselors and operations

Smart Documents

Prepare communications from approved context already in the workflow.

AI assistance can draft passages for letters, email, resources, and operational documents while templates and authorized CMA data provide structure.

Staff can spend more time checking the message and less time rebuilding routine language.
Administrators and analysts

Custom SQL report assistance

Draft a read-only report query from an everyday request.

In Custom SQL Reports, authorized users can describe the information they need and use AI Builder to draft a read-only query. They review the SQL, fields, and results before using the report.

Operational questions can move from a plain-language request to a validated CMA report.
System administrators

AI Builder

Give authors a starting point for supported templates and SQL.

Authorized authors can use AI Builder while editing Document Templates, supported SQL fields in Task Management, and matching SQL for Smart Notes rules. The generated content remains editable.

Authors preview and validate the result before it becomes part of the agency workflow.

Example scenarios

See four ways teams can use AI in CMA.

These examples show the operating pattern. Exact availability depends on configuration, permissions, connected services, and organizational policy.

When learning a workflow

A staff member needs to understand enrollment and prepare a useful note.

  1. The authorized user asks CMA Assistant to explain the configured lead to client enrollment flow for their agency.
  2. The Assistant checks related settings and available reference guidance, explains the supported relationships, and can include a workflow diagram.
  3. The user asks for a note draft summarizing the explanation, checks the details and any gaps, and saves through the normal CMA workflow.

During a client call

A counselor needs useful guidance during a client conversation.

  1. An eligible call produces a live transcript for an authorized counselor.
  2. The selected Call Assist prompt surfaces cues, checklist items, a summary, and a draft note.
  3. The counselor decides what is accurate, what belongs in the client record, and what action to take.

After client contact

A service team needs a consistent record and a clear next step.

  1. A configured Smart Notes rule matches an eligible call transcript and runs the approved prompt.
  2. CMA creates a Client Note under the selected Note Type and links it to the source Call Log.
  3. Staff compare material details and next steps with the transcript or recording, then correct the workflow or add clarification when needed.

When a task changes

A task reaches a point that calls for repeatable analysis.

  1. An administrator links an active AI prompt template to Task creation or completion for the appropriate agency workflow.
  2. CMA combines the template with permitted client context and runs it when that Task event occurs.
  3. The result is stored as an AI ClientMessage so authorized staff can review it and decide what happens next.

Responsible operation

Keep people responsible for every result.

AI assistance supports the agency’s process. People remain responsible for professional judgment, policy decisions, legal review, and the client record.

Human review remains part of the workflowA person checks each generated result and makes the final decision.

Configuration

Choose where assistance fits, which prompt applies, and what the workflow should produce.

Authorized access

Limit AI features and eligible context according to the permissions defined for the deployment.

Approved context

Use the records, templates, and content appropriate for the specific task.

Human review

Keep people responsible for accuracy, professional judgment, and the final action.

Consent and policy

Align recording and AI use with organizational policy and applicable disclosure and consent procedures.

Validation

Check generated notes, documents, template logic, queries, and reporting results before use.

Map the right starting point

Map AI to a workflow your team wants to improve.

See how AI assistance can fit into your agency’s workflow, from relevant context and review responsibility to the final CMA action.