If this project has been attempted a few times and been canned, chances are there's real business impact from collaborating on and delivering this item. The biggest thing I can recommend is to have a meeting and ask them but to also think about it in terms of a holistic tech picture. How does what they need serve to where we want to get.
Creating and maintaining confidence in a project is partially a matter of how much depth an audience has in a specific knowledge area and how well they know it. For example, I can communicate really novel AI designs since I have deep expertise in ML/AI models and have gained a lot of respect from building that knowledge out and sharing it with everyone. When I go to work, I can be like we can do this novel thing, it's really easy, let me show you how etc and then you implement it and you're done. Instance confidence. Furthermore, keeping them in the loop during the execution of the project and having broken it down into something that multiple teams can work on is also a great help. People who are skilled in servers for instance aren't always the best developers, but you might need a lot of developers for implementing the in-depth code and then higher level folks maybe one or two people for setting up an environment that can run that in production. It all depends on what you're interested in though and what you want to be an expert in, but always plan for an appropriate sized team.
Maybe establish partnerships with neighboring companies interested in sharing and see what good work comes of it. You never know what expertise can be brought to a problem unless you feel comfortable sharing it. The more the better to a point in my opinion.