About AmpAI¶

AmpAI is the engine that drives the conversational AI tools within Amperity, including the Customer Data Agent and the AI Assistants. AmpAI lets you ask and answer complex questions about your customers, with support for visualizations, and operates over the data in the database.

AmpAI supports customization through custom prompting capabilities. This enables organizations to encode specific business logic and defintions, tailoring AmpAI to align with their business requirements and ensuring consistency in results.

AmpAI offers enterprise-grade privacy and security and is built on the Azure OpenAI Service.

To enhance response accuracy, AmpAI leverages tenant-specific information, such as schema metadata, database field descriptions, and usage patterns to refine its understanding and to improve the relevance and precision of results.

AmpAI enforces user-level permissions, allowing granular control over access.

Note

AmpAI is powered by models that reside on Azure OpenAI Service. Review the AmpAI Privacy FAQ for information about how Amperity interacts with Azure OpenAI service.

AmpAI tools¶

AmpAI encompasses multiple tools, each designed for a different stage of your workflow. The Customer Data Agent is a conversational starting point — describe what you want to accomplish and it generates segments or journeys from scratch. The AI Assistants are embedded in individual editors for segments, journeys, and queries, where they help with detailed refinements, while the Consumption AI Assistant helps you understand how you are using Amps.

A typical workflow might start with the Customer Data Agent to quickly create a segment and journey, then use the Manual edit option to open the specialized editors where the AI Assistants can help with detailed adjustments.

Note

Custom prompts set in AmpAI apply to all AmpAI tools, including the Customer Data Agent and all AI Assistants.

Tool	Best for	Access point
Customer Data Agent	Starting from intent, creating segments and journeys from scratch, multi-step workflows	AmpAI button in sidebar
Segments AI Assistant	Fine-tuning existing segments, making precise adjustments within the segment editor	Within the Segments page
Journeys AI Assistant	Refining journey logic, adjusting channel configurations within the journey editor	Within the Journeys page
Queries AI Assistant	Writing and debugging SQL queries, data exploration	Within the Queries page
Consumption AI Assistant	Monitoring Amps consumption	Within the Amps page

Learn more about Customer Data Agent and AI Assistants.

Getting good results with AmpAI¶

To achieve optimal results when using AmpAI, follow these best practices for structuring your prompts:

Understand the question’s scope

Define the scope of your question to avoid ambiguous results. For example, specify the timeframe, customer segments, or metrics you are analyzing.
Avoid overloading

Focus on one primary question or task per prompt to ensure clarity and to avoid confusing results.

For example, instead of asking how the demographics of omnichannel customers compare to single-channel customers, ask a question first about omnichannel customer demographics, and then ask a second question about single-channel customer demographics.
Use consistent terminology

Stick to terminology used in your schema and business logic to align with how AmpAI understands your data.
Update the custom prompt often

The custom prompt is a powerful tool. Update the custom prompt whenever you get a result that does not align with the way your business views the world.

About the custom prompt¶

The custom prompt in AmpAI enables your brand to tailor responses to align with specific business logic. Custom prompts apply to all AmpAI tools, including the Customer Data Agent and AI Assistants, and can be updated to include:

Customer definitions, such as defining how your brand interprets retention metrics
Priority tables and fields, such as specifying priority for default tables and fields that are used with loyalty-related queries
Exclusions, such as automatically excluding employees or outliers from analysis
Default customer identifiers
Fiscal calendar information

While AmpAI works out of the box, updating the custom prompt is often necessary to keep AmpAI aligned with how your brand understands your customers. The custom prompt should be updated when an AmpAI response does not meet expectations.

Configure AmpAI¶

When your brand starts using AmpAI you should start with a list of 3-5 key customer questions that AmpAI should answer correctly. Test these questions, and then refine the custom prompt to ensure accurate and meaningful responses. This will help establish an effective first iteration of the custom prompt.

Custom prompt iteration¶

AmpAI supports two types of prompts for customization and iteration:

Production prompt

The production prompt applies universally to all users and queries within AmpAI. Any changes to this prompt should be deliberate and well-tested using the draft prompt.
Draft prompt

The draft prompt is a testing environment where changes can be made and evaluated without affecting all users. Changes to the draft prompt only apply to your specific instance of AmpAI and not to any other users.

Use the draft prompt to test modifications until the changes are ready for deployment.

Click “Revert prompt” to revert the draft prompt to the current version of the production prompt.

Click “Activate draft” to update the production prompt to match the current version of the draft prompt. This will make the updates available to all users.

Write an effective custom prompt¶

Follow these guidelines to create custom prompts that help AmpAI understand your brand’s specific terminology, business logic, and data structure.

Use database section headers¶

The custom prompt is set per tenant, but a tenant may have multiple databases. To define different behavior per database, use the database name as a section header. When AmpAI runs a session, it operates against a single active database. The system prompt instructs AmpAI to use the section matching the active database name for additional context.

If a tenant has rules for multiple databases, structure the prompt like this:

C360
Use Customer360 for all customer profile data.
"High-value customer" means lifetime revenue > $1,500.
Always exclude employees using Employee_Flag = 'N'.

Marketing_DB
Use Campaign_Customers for profile data.
"Engaged customer" means opened an email in the last 90 days.

If the tenant only has one database, the header is still recommended for clarity but is less critical.

Be specific with exact names¶

Reference exact field names, values, and table names whenever possible to ensure precise results.

Avoid vague references:

Bad: “Use the main customer table for profiles” Good: “Use Customer360 for all customer profile data (names, contact info, attributes)”

Bad: “Filter out internal users” Good: “Exclude records where email ends with @acmeretail.com, @acmeinternal.com, or @testvendor.com”

Map business terms to Amperity terms¶

Customers often use business terminology that differs from your schema column names and values. Define these mappings explicitly in the custom prompt.

For example, customers say “SKU” but the column is product_id. Customers say “online” but the field value is ecommerce.

When users say "SKU," "product code," or "item number," use the product_id
column from Unified_Itemized_Transactions.

When users say "online," "ecom," or "ecommerce," filter on
channel = 'ecommerce' in the TRANSACTIONS table.

When users say "store" or "in-store," filter on channel = 'retail'
in the TRANSACTIONS table.

Provide SQL patterns for complex logic¶

For reusable filters and complex business logic, give AmpAI a complete CTE it can copy and adapt.

WITH real_customers AS (
  SELECT AMPID
  FROM Customer360 c
  JOIN (
    SELECT AMPID, COUNT(txn_nbr) AS txn_count
    FROM TRANSACTIONS
    GROUP BY AMPID
    HAVING COUNT(txn_nbr) BETWEEN 1 AND 60
  ) t ON c.AMPID = t.AMPID
  WHERE c.Employee_Flag = 'N'
    AND c.email NOT LIKE '%@acmeretail.com'
    AND c.email NOT LIKE '%@acmeinternal.com'
    AND c.email NOT LIKE '%@testvendor.com'
)

Define business concepts¶

If your brand has specific definitions for retention, churn, high-value, or loyalty tiers, spell them out explicitly.

"High-value customer" means a customer with lifetime revenue > $1,500
(top 25% of active customers).

"Churning customer" means a customer with < 50% probability of returning
in the next 12 months AND has made at least one purchase in the past 2 years.

"Active customer" means at least 1 purchase in the last 365 days.

Loyalty tiers:
- Bronze: lifetime revenue $0-$250
- Silver: lifetime revenue $251-$1,000
- Gold: lifetime revenue $1,001-$2,500
- Platinum: lifetime revenue > $2,500

Specify default exclusions¶

If every query should filter out test accounts, employees, or bots, state it once in the custom prompt.

Always apply the following filters unless the user explicitly asks
for unfiltered results:
- Employee_Flag = 'N'
- Total transaction count between 1 and 60
- Exclude email domains: @acmeretail.com, @acmeinternal.com, @testvendor.com

Test custom prompts¶

Testing is not “ask a question and see if it looks right.” Use a structured approach to validate that your custom prompt produces the expected results.

Define test categories.

Before writing test cases, identify what your prompt is supposed to do.

Category	What to test	Example
Term mapping	Brand terminology resolves to correct columns/values	“Show me online orders” uses `channel = 'ecommerce'`
Default filters	Exclusions are applied automatically	Any customer count query excludes employees
Table routing	AmpAI uses the right table	Profile questions go to Customer360, not Unified_Coalesced
Business definitions	Concepts match brand meaning	“High-value customers” uses lifetime revenue > $1,500
Edge cases	Opt-out works, ambiguity is handled	“Show me ALL customers including employees” skips exclusion filter

Write test questions.

For each category, write 1-2 natural-language questions a real user would ask. Use the customer’s actual vocabulary, not Amperity terminology.
Run tests in draft mode.

Set your custom prompt as the draft prompt, open AmpAI in a new conversation, and ask each test question one at a time. For each response, check:
- Did it use the correct table?
- Did it use the correct column names?
- Did it apply the expected filters?
- Did the SQL match your expected pattern?
- If it ran a query, do the results make sense?

Record results.

Track your test results systematically to identify patterns and failures.

#	Category	Question	Expected	Actual	Pass?	Notes
1	Term mapping	“online orders”	channel = ‘ecommerce’	channel = ‘ecommerce’	Yes
2	Default filter	“how many customers”	Employee exclusion applied	No filter	No	Make exclusion rule stronger

Iterate.

For each failure, identify why AmpAI didn’t follow the instruction. Common reasons:
- Instruction was ambiguous (“use the customer table”—which one?)
- Instruction was buried in too much text (move critical rules to the top)
- Column/table name was wrong (check the actual schema)
- Conflicting instructions in the prompt
Make one change at a time to the draft prompt. Re-run the failing test case. Re-run passing test cases to check for regressions.
Promote to production.

Re-run the full test set one final time on the draft. Activate the draft to production. Have a second person run 2-3 test questions to sanity check.

Sample test prompts¶

Copy these into AmpAI with the draft prompt active. Adapt the expected results to your brand’s custom prompt.

Term mapping¶

"How many customers bought online last month?"
-> Verify: uses the correct channel column and value, not LIKE '%online%'

"What are the top 10 SKUs by revenue?"
-> Verify: uses product_id, not a literal column called "SKU"

"Show me in-store revenue by month for the last year"
-> Verify: uses channel = 'retail' (or whatever mapping you defined)

Default exclusion filters¶

"How many active customers do we have?"
-> Verify: employee exclusion, email domain exclusion, and transaction
   count filters are all applied

"What is our average customer lifetime value?"
-> Verify: same exclusions applied even though the question doesn't
   mention filtering

"Show me ALL records in the Customer360 table with no filters"
-> Verify: exclusions are NOT applied (user explicitly asked for everything)

"How many total rows are in Customer360?"
-> Verify: no exclusions — this is a raw count question, not a "customer" question

Table routing¶

"What is the average age of our customers?"
-> Verify: queries Customer360, not Unified_Transactions or Unified_Coalesced

"What was total revenue last quarter?"
-> Verify: queries TRANSACTIONS or Unified_Transactions, not Customer360

"Show me customer names alongside their last purchase date"
-> Verify: joins Customer360 (names) to the transaction table (purchase date)
   on the correct ID

Business definitions¶

"How many high-value customers do we have?"
-> Verify: uses your exact definition (e.g., lifetime revenue > $1,500),
   not an invented threshold

"Break down customers by loyalty tier"
-> Verify: uses your tier ranges (Bronze/Silver/Gold/Platinum),
   not made-up buckets

"How many customers are at risk of churning?"
-> Verify: uses your churn definition, not a generic "hasn't purchased in 90 days"

Stress tests¶

"How many high-value online customers churned last quarter?"
-> Verify: combines business definition + term mapping + time filter
   + default exclusions all correctly

"Compare revenue from loyal vs churning customers"
-> Verify: uses your definitions for both terms in the same query

"Who are our top 100 customers?"
-> Verify: uses your definition of "top" (by spend? by frequency?)
   and applies exclusions

Segment and journey prompts¶

"Create a segment of high-value customers at risk of churning"
-> Verify: delegates to the Segment Agent and the resulting segment
   uses your definitions

"Build me a segment of customers who bought online in the last 30 days"
-> Verify: segment uses the correct channel mapping from your prompt

"Create a win-back journey for churning platinum customers"
-> Verify: creates a segment using your churn + tier definitions,
   then creates the journey

Common prompt patterns by industry¶

Use these patterns as a starting point when writing custom prompts for your industry.

Retail and e-commerce¶

Map “online” / “ecom” / “web” to channel value
Define loyalty tiers with spend thresholds
Exclude employee purchases and internal accounts
Map “SKU” / “item” / “product code” to product_id

Financial services¶

Define “active account” vs “dormant account”
Specify which transaction types count as revenue
Exclude internal/test accounts by account type

Hospitality¶

Map “stays” / “visits” / “bookings” to transaction table
Define “loyalty member” vs “guest”
Specify how to calculate LTV (room revenue only vs total spend)

Custom prompt limitations¶

The custom prompt cannot:

Change the AI’s tools or capabilities
Override core SQL syntax rules (Presto SQL)
Make AmpAI access tables outside the active database
Configure destinations, campaigns, or orchestrations
For segment/journey creation, the custom prompt provides context but the Segment Agent and Journey Agent have their own specialized logic

Permissions and policies¶

AmpAI permissions are controlled at the user level, allowing User Administrators the ability to grant access to AmpAI for individual users.

AmpAI has the following user-level policy options:

Restrict AmpAI access

Prevents users from accessing the AmpAI page.
Restrict Queries AI Assistant access

Prevents users from accessing the AmpAI Assistant from within the Queries page.
Restrict Segments AI Assistant access

Prevents users from accessing the AmpAI Assistant from within the Segments page.
Allow prompt administration

Allows users to update the custom prompt. Datagrid Operators and Datagrid Administrators always have the ability to modify prompts.

Disable AmpAI features¶

The Customer Data Agent and the AmpAI Assistants can be disabled for all users. Open the Settings page, select the AmpAI tab, and then click Disable AmpAI features.

Audit conversations¶

Customer Data Agent and AmpAI Assistant conversations can be audited by users assigned the Datagrid Operator and Datagrid Administrator policies from the Settings page. The AmpAI tab on the Settings page logs the questions that are asked to the Customer Data Agent and the AmpAI Assistants under AI Conversations.

The Activity log tab on the Settings page logs when AmpAI Assistant questions are asked using the “amperity.query.exec/sampled” action.

Choose topic collections

About AmpAI¶

AmpAI tools¶

Getting good results with AmpAI¶

About the custom prompt¶

Configure AmpAI¶

Custom prompt iteration¶

Write an effective custom prompt¶

Use database section headers¶

Be specific with exact names¶

Map business terms to Amperity terms¶

Provide SQL patterns for complex logic¶

Define business concepts¶

Specify default exclusions¶

Test custom prompts¶

Sample test prompts¶

Term mapping¶

Default exclusion filters¶

Table routing¶

Business definitions¶

Stress tests¶

Segment and journey prompts¶

Common prompt patterns by industry¶

Retail and e-commerce¶

Financial services¶

Hospitality¶

Custom prompt limitations¶

Permissions and policies¶

Disable AmpAI features¶

Audit conversations¶