Has anyone implemented a knowledgebase tool and repository with AI in the past 2 years?  Interested to connect with experience and solutions. 

We've developed a tool based on GenAI that integrates with our Azure Devops Wiki (where we currently document our systems) to Google Gemini. Google Gemini automatically generates systems documentation artifacts based on the wiki's data and user inputs, such as texts, documents, PDFs, images on a multimodal solution that enables our users to be more efficient and productive.

Now we are implementing RAG Proof of Concepts with our data, trying to integrate our legacy system documentation to give our users a tool that helps them to make better decisions. Our next step is to put a layer of a vector database so that we can improve our LLM answers. It's been quite an adventure bringing all these tools altogether, but now our measurements show that users who are using the tool in a good way are already seeing excellent results.

We have utilized the Salesforce core platform to store knowledge base articles, enabling healthcare insurers to easily access this information.

We have been developing various RAG-based solutions for the last 15 months or so based on existing software documentation, regulatory data, service tickets, etc...