I'm looking into setting up an AI-powered document repository for long-term scalability. I have a lot to learn and understand, but I'm trying to move in that direction now. The goal is to efficiently store and retrieve project estimate-related documents, so I can quickly pull relevant information when needed-whether it's technical specs, cost estimates, cost indices/databases, quotes, procedures, or past project details. I want to pair this with a chatbot to answer text questions on demand.

For those who have worked on something similar, what approaches have worked well for you? Any key challenges or lessons learned that you'd recommend considering?

I'm looking into setting up an AI-powered document repository for long-term scalability. I have a lot to learn and understand, but I'm trying to move in that direction now. The goal is to efficiently store and retrieve project estimate-related documents, so I can quickly pull relevant information when needed-whether it's technical specs, cost estimates, cost indices/databases, quotes, procedures, or past project details. I want to pair this with a chatbot to answer text questions on demand.

For those who have worked on something similar, what approaches have worked well for you? Any key challenges or lessons learned that you'd recommend considering?

I have been looking into doing something similar. Have you tried using Microsoft Copilot or Google Gemini on you document repository in Sharepoint/Google Drive? That might give you at least some of what you are looking for (and I assume what Craig Stadnyk is talking about).

You could also try https://notebooklm.google.com/ for smaller document sets.

For a more complete solution, there are some off-the-shelf services that could be repurposed for this I think (https://www.docugami.com/, https://www.graft.com/). Then there are several combinations of commercial and opensource software you can use to build something yourself.

This is a good reminder for me to try to put together something workable myself. I'll see what I can come up with.

I'm looking into setting up an AI-powered document repository for long-term scalability. I have a lot to learn and understand, but I'm trying to move in that direction now. The goal is to efficiently store and retrieve project estimate-related documents, so I can quickly pull relevant information when needed-whether it's technical specs, cost estimates, cost indices/databases, quotes, procedures, or past project details. I want to pair this with a chatbot to answer text questions on demand.

For those who have worked on something similar, what approaches have worked well for you? Any key challenges or lessons learned that you'd recommend considering?

Thanks, Konner.  I haven't tried anything yet.  Connecting to my company's SharePoint isn't allowed, so I was hoping to start with something independent of that.  I'm hoping to arm myself with enough knowledge that I can propose exactly what I want to build.  With things developing so quickly in the AI world, I'm hesitant to get started on something knowing something new and shinier may be just around the corner.

I have been looking into doing something similar. Have you tried using Microsoft Copilot or Google Gemini on you document repository in Sharepoint/Google Drive? That might give you at least some of what you are looking for (and I assume what Craig Stadnyk is talking about).

You could also try https://notebooklm.google.com/ for smaller document sets.

For a more complete solution, there are some off-the-shelf services that could be repurposed for this I think (https://www.docugami.com/, https://www.graft.com/). Then there are several combinations of commercial and opensource software you can use to build something yourself.

This is a good reminder for me to try to put together something workable myself. I'll see what I can come up with.

I'm looking into setting up an AI-powered document repository for long-term scalability. I have a lot to learn and understand, but I'm trying to move in that direction now. The goal is to efficiently store and retrieve project estimate-related documents, so I can quickly pull relevant information when needed-whether it's technical specs, cost estimates, cost indices/databases, quotes, procedures, or past project details. I want to pair this with a chatbot to answer text questions on demand.

For those who have worked on something similar, what approaches have worked well for you? Any key challenges or lessons learned that you'd recommend considering?

I'm looking into setting up an AI-powered document repository for long-term scalability. I have a lot to learn and understand, but I'm trying to move in that direction now. The goal is to efficiently store and retrieve project estimate-related documents, so I can quickly pull relevant information when needed-whether it's technical specs, cost estimates, cost indices/databases, quotes, procedures, or past project details. I want to pair this with a chatbot to answer text questions on demand.

For those who have worked on something similar, what approaches have worked well for you? Any key challenges or lessons learned that you'd recommend considering?