What we heard at InnoTrans

We spent three days at InnoTrans in Berlin this September, in meetings with clients and partners and walking up to stands to talk to whoever was there. It is the world’s largest rail trade fair, with more than 130 vehicles on display across over 40 halls, of which we managed about three. Whatever a conversation started on, it usually came round to the same subject: how to use AI on complex bids, and why it is proving harder than people expected.

From the field · Bid partnerships, rail and infrastructure · 8 October 2026

Copilot, and not much else

Most of the companies we spoke to have no plan for AI in complex bidding beyond giving everyone Copilot as a personal productivity tool. Copilot is good at that, and bid writers should use it, but a large tender asks for something different. The whole team needs one shared reading of the brief, the evidence for each question has to be found and checked, and the answers have to be tested against what the evaluator is actually scoring. A dozen people each using their own assistant does not give you any of that.

Through 20 years of winning bids for our clients, the bidding and engineering team at Blue Edge know more about this subject than any other company we have come across, and we have built that knowledge into a platform. Blue Edge AI does the structure, analysis, planning and review of a complex bid, and leaves the thinking and the writing with the bid team, where they belong. We wrote about why in-house builds have struggled to cover the same ground in The in-house build.

The people who want it most are often held back

The bid directors we met are working hard to find ways of using AI to help them win, and many of them are held back by their own company’s IT strategy, which was rarely written with bidding in mind. The tender still has to go in on time, so some are getting round it by bringing in Blue Edge AI for a single bid and paying for it from that bid’s budget. It does not need an enterprise roll-out to be useful on one tender, and because it is available through the Microsoft Marketplace it is easier for IT and procurement to say yes.

The firms that are furthest ahead are not the ones we expected. Medium-sized companies are making far better use of AI and of bid process than the large corporates, largely because the person who runs the bids can make the decision, so there is very little between trying a tool and using it on a live tender. With teams of five to thirty work-winning people we deploy the platform, support a few early users through onboarding, and the team carries on from there.

Whatever the size of the firm, it is early days, and choosing the right partner matters. Getting value from AI in bidding is as much about people and process as about the technology. Our developers are led by the same bid professionals who have spent 20 years winning bids for our clients, and alongside the platform we provide bid support on live tenders and coaching for bid leaders and their teams, including orals and the Q&A sessions where the evaluation panel is in the room.

And the rest of the show

Engineers are great to work with, because they are problem solvers to the core, and the bid directors running £7bn bids are an impressive group of people. Three days of meeting people face to face was a reminder of how much better that is than Teams. The trains were cool too, although none of us is a spotter yet.

If your team is weighing up what to do about AI in bidding, we can show you the platform on a tender you have already bid, using your own documents, in person or online. That is usually the quickest way to see whether it would help on the next one.