Create Document QA Bot with Robomotion RPA
In this video we are going to build a Document QA Bot Flow with Robomotion RPA. We're using Google Sheets to store our Q&A pairs, making it easy to manage and update. On top of this, we've built a Robomotion Assistant Bot that interacts with the database to provide answers. For better matching, we're utilizing OpenAI's Embeddings API, and for more contextual responses, we're integrating ChatGPT.
If you find this content useful, feel free to hit the like button. For more technical walkthroughs and tutorials, don't forget to subscribe.
Join our Discord 👉 https://community.robomotion.io
Here is the shared flow:
https://learn.robomotion.io/designer/...
Tips:
To improve the precision and context-relevance of ChatGPT's output, consider setting the temperature to either 0.2 or 0. For more efficient data handling, it's advisable to break down the QA database into discrete questions and answers, adding them to a CSV file one row at a time. You could use the "Create Embeddings" node instead of the "Create Batch Embeddings" node for this task, and then append the resulting vector to a new column. This method is more effective than merging text and segmenting it across unrelated sections.
#Robomotion #RPA #Chatbot #OpenAI #GoogleSheets #Automation