AI Productivity Helpers

From Grooper Wiki

This article is about the current version of Grooper.

Note that some content may still need to be updated.

2025


There are several productivity features in Grooper that use large language models (LLMs) to simplify a users work in Grooper. LLMs (such as OpenAI's GPT models) allow users to tell Grooper what to do using a natural language text prompt. There are currently six (6) of these features in Grooper version 2025.

  • BE AWARE! For each of these features to work, users must first have an LLM Connector configured at the Grooper Root node and a "Tools Chat Model" configured.

You may download the ZIP(s) below and upload it into your own Grooper environment (version 2025). The first contains one or more Batches of sample documents. The second contains one or more Projects with resources used in examples throughout this article.

List of AI productivity tools:

  • AI Generated (Schema Importer)
  • AI Query Helper
  • AI Expression Helper
  • AI Database Lookup Helper
  • AI XML Transform Helper
  • AI Helper


Data Model Schema Importer: AI Generated

Grooper's "Schema Importers" generate data_table Data Models from various sources, such as database schemas or XML schemas. The "AI Generated" importer generates a Data Model from a natural language prompt using an LLM.

This Data Model was generated using the following simple prompt:

Create a Data Model for invoice processing.

AI Generated created a simple Data Model for processing invoices. However, more specific prompts will yield more specific results. If you tell the LLM more specifically what fields you want to add, it will do so. You can even paste a document's full OCR text into the prompt and tell it "Make a Data Model for this document".

Furthermore, when it determines a Data Field or Data Column should be a datetime or numeric type (double, decimal, int16, etc.), it will adjust its Value Type configuration accordingly.

The AI Generated importer will even add brief descriptions to the Data Elements it creates in some cases.



To use the AI Generated importer:

  1. Right-click a Data Model in the node tree.
  2. Select the "Import Schema..." command.
  3. Expand the "Source" dropdown (Press the "☰" button).
  4. Select "AI Generated".
  5. Open the "Instructions" editor (Press the"..." button).
  6. In the Instructions editor, enter a prompt for what you want to create.
    • Example: Create a Data Model for invoice processing.
    • Example: Create a Data Model for processing student transcripts.
    • Example: Create a Data Model for this document: (Then paste a document's text into the Instructions editor)
  7. When finished, press "OK" to exit the Instructions editor.
  8. Press the "Execute" button in the "Import Schema" window to generate the Data Model.

AI Query Helper: Get help on the Search page

Example prompt. The AI Query Helper also lists all fields in the search index.

The Grooper Search Page is a powerful document search and retrieval tool. Part of its power lies in the specialized syntaxes used in the search parameters. This can be challenging for novice users.

What if you need help making a search query? The "AI Query Helper" can help!

The AI Query Helper allows any user to return a document set using a simple text prompt. This helper is accessed in the Search Page using the "AI Query Helper" button (cognition).

Example Prompt: Find invoices greater than $10,000 from before 2020. Only show me the invoice number, invoice date, and invoice amount. Sort by the largest invoice amount.

The AI Query Helper hands the prompt and a description of the fields in the search index to a large language model (LLM) AI. With that in hand, it crafts the search query for you and fills in the Query Editor. All you have to do from there is hit the Search button!

To use the AI Query Helper:

  1. Go to the Search page.
  2. Select the search index you want to query.
  3. Press the "AI Query Helper" button (cognition) to bring up the helper window.
  4. Write what you want to search for in plain language.
    • The AI Query Helper editor will also show you a list of fields in the search index. This can help you get a better idea of what fields can be referenced in the search query.
    • Most typically the helper will fill in the Search and Filter parameters using your prompt. However, if you include language like "Only show me fields X, Y and Z" and "Order results by field X." It will fill in the Select and Order By parameters too.
  5. Press the "OK" button when finished.
  6. The helper will fill in the search query parameters using the prompt.
  7. Press the "Search" button (search) to execute the search query.

AI Expression Helper: Get help making Pattern Match regex

Regular expression (regex) patterns are powerful tools used to find patterns in text. Grooper's Pattern Match extractor uses regular expressions to match text and return them as results in a result set. Pattern Match extractors are often key in one component of a users document processing solution (be it in document separation, classification, data extraction or some other area).

What if you don't know how to write a regex pattern? The "AI Expression Helper" is here to help!

The AI Expression Helper can be accessed in the Pattern Match editor. Tell it what kind of pattern you want using simple natural language.

Example: Generate a pattern that matches common date formats.
Example Result: (\b(?<Month>0?[1-9]|1[0-2])[\/\-](?<Day>0?[1-9]|[12][0-9]|3[01])[\/\-](?<Year>\d{4})\b
  • AI responses are varied! Because this helper uses generative AI, the expression will likely be someone different even if you ask the same question. Ask multiple times to try out variations on the same pattern!

The helper automatically configures the Pattern Match editor for you. Even if the pattern isn't exactly what you want, it a great starting point to get extracting results quickly.


To use the AI Expression Helper:

  1. Open a "Pattern Match" extractor editor (Press the "..." button).
  2. Press the "AI Expression Helper" button (cognition).
  3. In the editor, enter a prompt for the regular expression pattern you want.
    • Example: Generate a pattern that matches common date formats.
    • Example: Show me the money. (It will match currency values)
  4. Press the "OK" button when finished.
  5. The helper will fill in the expression.
  6. When satisfied with the expression, press "OK" to exit the Pattern Match extractor editor.

AI DB Lookup Helper: Get help writing SQL queries for Database Lookup

This query was generated by AI DB Lookup Helper.

"Lookup Specifications" (or "lookups") are used in Grooper to take data extracted in a document and use it to find data in external sources like in databases, XML files, or web services. This can help validate document data and populate additional data not present on the document.

"Database Lookup" uses a SQL query to look up data in SQL database tables. Grooper "lookup fields" are referenced in the query's WHERE clause to return values to "target fields" in the SELECT clause.

What if you don't know how to write a SQL query? "AI DB Lookup Helper" is here for you!

  • Be aware! You should always verify the query is correct and does not unintentionally alter your database before putting it in production. It is possible for an AI to generate queries with statements that make changes to a table (UPDATE, INSERT, DELETE). If the Data Connection user has rights to alter the database, the Database Lookup will alter your database.


To use the "AI DB Lookup Helper":

  1. Ensure you already have a database Data Connection in your package_2 Project.
  2. Go to the Data Element where Database Lookup is configured.
    • Database Lookups can be added to data_table Data Models, insert_page_break Data Sections, or table Data Tables.
  3. Under the "Behavior" properties, open the "Lookups" editor (Press the "..." button).
  4. If not done so already, add a Database Lookup specification. Otherwise, select a Database Lookup specification from the list.
  5. If not done so already, configure the "Database Connection" property by selecting the database Data Connection in your package_2 Project.
  6. Open the "SQL Query" editor (Press the "..." button).
  7. Press the "AI DB Lookup Helper" button (cognition).
  8. Write what you want the SQL query to do.
    • Example: Make a query that looks up an order number given an order date, order amount and a customer's name.
    • Example: Verify the order amount matches what is in the database for the order number.
      • AI generated validation queries can be tricky. Always ensure "Field Population" is set to "None" when using validation-only lookups. You may also need to set the "Miss Disposition" to "Flag" to properly flag the fields as errored.
  9. The AI will generate the SQL query for you. Edit it in the "SQL Query" editor as needed.
  10. Press "OK" when finished.
  11. Finish editing the rest of the Database Lookup properties, if needed.
  12. Press the "OK" button in the Lookups editor when finished.
  13. Press the Save button (save) when to save changes to the Data Element.

AI XML Transform Helper: Get help making XSLT transforms

XML Transform is an activity that transforms XML data to a new file using an XSLT (eXtensible Stylesheet Language Transformations) stylesheet. XML Transform can be used to:

  • Convert XML data into different XML schemas for interoperability.
  • Generate human-readable formats (such as HTML) from XML for reporting or review.
  • Prepare XML for downstream integration, export, or ingestion by other systems (including AI knowledge bases).
  • Normalize or restructure XML from various sources to a standard schema.

FYI

The XML source can be:

  • A document's extracted Data Model (serialized as XML)
  • A document's primary attachment file (if it is an XML file)
  • An XML file attached to the document by a custom Activity or Command.


To use XML Transform, you need an XSLT stylesheet. What if you don't know how to write XSLT code? The "AI XML Transform Helper" is here to help!

To use the AI XML Transform Helper:

  1. In the Grooper node tree, navigate to the XML Transform step in a Batch Process.
  2. Configure the "Source" property, as needed.
  3. Go to the "XSLT Editor" tab.
  4. Press the "AI XML Transform Helper" button (cognition).
  5. Write instructions on how you want the XML to be altered.
    • Example: Transform this XML into an HTML report. (This would transform source data into a more human-readable document by transforming the XML data to HTML tags)
    • Example: Strip out all the attributes except "Name". (This would remove all attributes from the Data Model data serialized as XML except "Name". This make a leaner XML file, getting rid of attributes like "Confidence", "Page" and "Location")
  6. Press "OK" to generate the XSLT.
  7. The helper will generate the XSLT and insert it into the "XSLT Code" window.
    • You can test the transform with the "Test" button (play_circle).
    • You can edit the XSLT in the "XSLT Code" window if needed.
  8. When satisfied, press the "Save" button (save).

AI Helper: Get help all over the place in Grooper

Just look for the "AI Helper" button.

The "AI Helper" is a general purpose helper available to many editors in Grooper. You can access the AI Helper in any of Grooper's "Code Editors". These editors show up when editing many different properties in Grooper (and not just when you're editing code expressions!). Just look for the "AI Helper" button at the top of the editor.

"AI Helper" uses include:

  • Get help creating list entries for List Match.
  • Get help with Grooper code expressions (Default Value, Calculated Value etc).
    • Calculated Value Example: Make an expression to calculate the total amount due from the all rows in the Invoice Line Items table.
    • Calculated Value Example: Make sure this number is what it's supposed to be.
      Author's note: I asked the AI to see if it would make the right expression to ensure a line total value in an invoice line items table equaled the unit price multiplied by the quantity. Even with that little context, the AI came up with the right expression (However, I'm unsure how an AI would respond to that prompt in all cases).
  • Get help making instructions for AI Extract.
  • Get help making a Description for a Grooper node.
Retrieved from "https://wiki.grooper.com/index.php?title=AI_Productivity_Helpers&oldid=29436"