Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
With Qlik Talend Cloud, Data Integration pipelines can be created to curate data for AI Retrieval Augmented Generation (RAG) solutions. You can reduce time to value for gaining insight into your enterprise data using Generative AI by leveraging Qlik Answers. QTC will automate loading structured and unstructured data sources to create knowledge base assistants that users can prompt for questions and retrieve answers. QTC will simplify the complexity of transforming data by chunking data for vector embedding and building the context for prompts, within an easily deployable conversational agent.
Introducing Qlik Answers
Qlik Answers is a plug-and-play, generative AI-powered knowledge assistant that provides users with the ability to gain answers to questions from the user’s sourced content. Generative AI is used to personalize answers to prompted questions of the indexed data sets within the user created knowledge base. Qlik Answers is available in the Qlik Talend Cloud and can be used with Data Integration pipelines to provide an end-to-end solution for your RAG use cases.
Qlik Talend Cloud brings it all together
The capabilities of QTC allow you to use automation to create a data pipeline that can ingest data from any supported source into a target for integration with Qlik Answers.
Using QTC we can show an example of loading structured data to a Databricks target and leverage the data for Qlik Answers (Hawaii resorts structured location data set used for source data.)
Setting up and running Qlik Talend Cloud Services
Create a Databricks target data pipeline to onboard the Hawaii resorts source data into a Databricks Delta table.
Create an AI-ready transformation off the Storage task within the QTC data pipeline to ingest data for Qlik Answers knowledge base utilizing an AWS S3 object store.
Add the Hawaii location reserve dataset to the transformation for preparation and loading.
The complete data pipeline is shown below.
The reserve data will be loaded in the AWS S3 location after the data pipeline is completed.
Upload the unstructured data set to the same AWS S3 bucket location. (Hawaii location Brochure unstructured pdf file.)
Within QTC choose Analytics tile and create a Knowledge Base with the files in the AWS S3 location.
Index the source files and create the Qlik Assistant
Use the assistant to ask questions on the indexed data sets. The following screenshots demonstrate the value of having that data in Qlik Answers (or other Retrieval Augmented Generation (RAG) solutions) by showing end users using the assistant to answer questions using that data. Of course, QTC Pipelines will keep that data up to date, so the answers will remain as valuable in the future as they are today.
Example 1 Chat with the chat bot.
Showing the unstructured source data used in the chat bot.
Example 2 Chat with the chat bot.
Showing the structured source data used in the chat bot.
Conclusion
With Qlik Talend Cloud, data pipelines can be used to create a no-code Retrieval-Augmented Generation (RAG) solution for data within your organization. QTC data pipelines will automatically ingest data that is structured into a target location that can be combined with unstructured data for use within Qlik Answers. Qlik Answers will simplify leveraging GenAI with the creation of a chatbot which can be used to directly prompt curated data sets for answers.
AI-Ready tasks are currently in private-preview in Qlik Talend Cloud. For more information or to take part in the private preview contact your Qlik account representative.