The Generate Mock Data workflow generates (a) structured records according to a JSON schema uploaded or defined by the user, and (b) synthetic texts in a user-defined format aligned with each record.
Select the View example outputs tab (in app) or navigate to example_outputs/generate_mock_data (on GitHub) for examples.
- [Input] A JSON file containing the JSON schema with which to generate output records (optional).
- [Process] The user edits the uploaded JSON schema or creates one interactively.
- [AI Calls] The system uses generative AI to create a dataset of mock records following this JSON schema.
- [AI Calls] The system uses generative AI to create a dataset of mock texts based on these mock records.
- [Output] A JSON schema defining structured data records, a dataset of mock records following this schema, and a dataset of mock texts following these records.
- The input schema, if provided, should be a JSON file conforming to the JSON schema standard and following the restrictions of the OpenAI Structured Outputs API.
Generate Mock Data can be used to create mock data for demonstration or evaluation of any other workflow.
Mock data is particularly helpful when working in sensitive domains and/or with personally identifiable information (PII).
The task for this tutorial is creating a mock case dataset of customer complaints, where each case record describes an identified individual and the nature of their complaint regarding a specific product. This is a useful proxy for any individual-level case data or "microdata" where the privacy of data subjects must be respected.
From the Generate Mock Data homepage in a running instance of Intelligence Toolkit, select Prepare data schema.
In the top left, you have the option to upload an existing JSON schema. This is helpful when the data schema represents an existing data standard or has been saved from previous use of the workflow. In this tutorial though, we will begin by defining a new schema using the Edit data schema form. Any changes made using this form will be reflected in the Preview pane to the right.
The initial JSON schema contains some boilerplate metadata fields representing best practices for schema design. The metadata fields are as follows:
$schema: Indicates the version of the json-schema standard that the current schema follows. Leave this field as it is.title: A short title indicating the kind of data that the schema represents.description: A longer description of the kind of data that the schema represents.records: The collection of records represented by the schema.
Try editing some of these metadata fields now, and see them reflected in the Preview of the JSON schema to the right. In particular, set the title field to Customer Complaints.
The schema in progress is validated after every change, with the message Schema is valid confirming that the current schema conforms to the standard specified in the $schema field.
Try downloading an edited schema using the download button, uploading it via the Upload schema control, then continuing as below.
Let's now specify the type of records represented by the schema by renaming records to complaint_records in the form to the left.
Next, we need to add fields to the objects of complaint_records for each attribute of the records we want to create.
Using the controls under Add field to complaint_records, press the str button to add a string (i.e., text) field. This field appears as the level 1 string string_1 (level 1 because the field is nested one level down from the complaint_records array at the top level, i.e., level 0). Edit the text label from string_1 to name.
As further string fields within complaint_records, now add:
streetandcityas string fieldsageas a numeric field using thenumbuttonemailas string fieldprice_issue,quality_issue,service_issue,delivery_issue,description_issueas boolean (true/false) fields using theboolbutton
Note that all new fields have the Required? checkbox checked by default, placing all field names in the required field of the object. This is a requirement for the OpenAI Structured Outputs API, which we'll later use to generate mock data that follows the schema. Similarly, all objects must also have additionalProperties set to false, so the Additional? checkbox is left unchecked by default.
Next, we want to add a product_code string field, but limit the possible values of the field to a predefined list called an "enumeration". Do this by checking the Enum? checkbox and observing the default values A, B, and C added to the enumeration. These values can be edited, deleted, and expanded as desired. For this tutorial, simply add further enum values alphabetically from D to H.
Note that boolean attributes of the record could also have been created using the str[] button to create a string array, checking Enum?, and specifying price_issue, quality_issue, service_issue, delivery_issue, description_issue as possible values. However, by using independent boolean fields we simplify the overall record structure and avoid the challenges of nested arrays in the final data object.
Finally, add a quarter string indicating in which calendar quarter the complaint was made. In the description, you can add a hint about the field structure and contents, e.g., "The quarter in which the complaint was made (since 2020-Q1)".
The schema is now complete and can be downloaded using the Download complaint_records_[schema].json button below the schema preview. Selecting the Sample object tab shows a minimal JSON object confirming to the schema.
Navigate to the Generate mock records tab to generate a mock dataset conforming to the Customer Complaints schema.
You will see that Primary record array has already been set to complaint_records, since this is the only array field in the schema. In the presence of multiple arrays, select the one that represents the primary record type whose records should be counted towards the Total records to generate target.
All Data configuration controls are as follows:
Records per batch: How many records to generate in a single LLM callParallel batches: In a single iteraion, how many batches to generate via parallel LLM callsTotal records to generate: How many records to generate. Must be a multiple ofRecords per batchxParallel batchesDuplicate records per batch: Within each batch, how many records should be near-duplicates of a seed record randomly selected from existing recordsRelated records per batch: Within each batch, how many records should appear closely related to (but not the same as) a seed record randomly selected from existing recordsAI data generation guidance: Guidance to the generative AI model about how mock data should be generated (e.g., targeted a specific region, time period, industry, etc.)
Press Generate mock records to generate mock data records according to the schema and configuration. Each record array in the JSON schema will be converted into CSV format and shown in its own tab (in this example, there will be just a single tab for complaint_records). Both the JSON object for the entire dataset and CSV files for each record array may be downloaded using the buttons provided.
Press the Download complaint_records_[mock_records].csv button to download structured record data for the next step.
Go to the Generate mock texts, select Browse files, and upload the complaint_records_[mock_records].csv file created in the previous step.
Expand File options and set Maximum rows to process (0 = all) to 50 for test purposes.
Under AI text generation guidance, enter any guidance to the generative AI model about how to style and format the text, e.g., in a standard document type for the relevant domain.
Given the domain of the example, try Generate output texts as transcripts of customer complaint calls.
Temperature is a number between 0 and 1 that affects the variation in AI outputs. Lower temperatures mean less variation, and AI outputs will be more accurate and deterministic. Higher temperatures will result in more diverse outputs.
Press Generate mock texts to see the mock texts for the given guidance and temperature generated on the right under Generated texts.