Is anyone using AI to aid in the generation of RFPs?

Like a few of the respondents, we use an LLM to respond to the thousands of RFPs we receive each year. This private implementation has had a very positive impact on our business. The significant investment in our robust Vector DB has made the difference. How many RFPs do you generate a year? 

I like the suggestion about loading up responses to a private model (You wouldn't want to just upload your RFPs to a public AI platform for obvious reasons).   This is probably the easiest way to get started.

Additionally I think one of the most interesting developing uses of GenAI for this purpose would be Retrieval Augmented Generation (RAG).  The possibilities with this are many fold and we have been helping clients explore the best uses for them in their enterprises.  You can have a repository of RFP's that it can learn on, this works best if you have plenty of RFP's to load up your own and publicly available ones (I imagine you probably have many of these to use).  

With RAG you can tune your prompts and responses and help it learn.  This is worth investigating for this and other use cases in your organization.  It is not as easy as you need to understand code a bit but it can be made with an interface to make it usable for the rest of us mortals.

Yes. We've loaded all our RFP responses into our private GPT model which we use as a starting point. This helps streamline the response process. The other thing we do for public/government RFPs that we didn't win, is to load the responses of the winning proposal, obtained via a public records request, and compare the two responses for insights.

Yes, it saves time by generating standard content that an analyst can then verify, supplement and refine.  

Hello,
it is one of the use case under evaluation but I can be more detailed as soon you share your needs.

Cheers.