AI in innovation management: Where it’s already in use and what’s yet to come

Artificial intelligence has already fully arrived in innovation management at neosfer. Whatever we do in terms of AI, it’s always about how the technology can make our daily work easier, so that we can act faster, more focused and more strategically. A very recent, simple example from day-to-day innovation at neosfer: as part of our venture building innovation process, we recently collected 120 ideas. Instead of sorting them manually, Google Gemini filtered out the redundancies and helped us cluster them. This was very efficient and gave us more time for actual thinking.

Things are also changing rapidly in the area of prototyping. Every MVP used to be built be our developers. However, today project managers can also do this themselves, albeit not at the same level as the devs. At neosfer, we now also build landing pages directly using no-code tools. Of course, this division would be out of the question for real products, but for initial versions and text pages with a clear and simple structure, it is a great relief. It also speeds up our venture building process because it prevents some of the innovation backlogs. At the same time, it allows our tech teams to focus more on more complex topics. This is where I see the biggest benefits of AI: it helps us to allocate our resources more sensibly.

Other areas in which we use AI include writing job advertisements, preparing foreign-language case studies, converting text into presentations, complex formulas or, in the content area, things like editing short videos from long videos, image processing and much more. It has long been standard practice for us to automatically transcribe and summarize meetings. What I personally also use a lot is to have research papers converted into a podcast so that I can listen to them on my way to work. This gives me a good overview and saves me a lot of time. And if I need even more details, I can go deeper. Many things that used to be tedious are now much quicker.

We are, however, also noticing something: With increasing variety of tools the risk of teams losing themselves in this veritable jungle of software also rises. Some people use Notion, others Confluence, others work with Trello. This makes collaboration complex and difficult to understand, especially when the composition of teams changes. That’s why we document our experiences with various AI tools and lessons learned directly in our Confluence wiki.

In the long term, I would like to see a new form of orchestration: when starting a project, all the necessary processes should already be running automatically in the background. The documentation is triggered, an MVP is created, everything is neatly stored in Confluence and at the end there is a finished presentation for management. And while all this is happening, a legal agent runs along to check whether regulatory requirements are being met, and a tech assistant also checks the MVP again. AI agents are not supposed to take over everything, but rather establish connections and distribute tasks intelligently. That would be a real game changer for innovation work.

One thing is certain: despite all the progress, we humans make the decisions. AI can support, prepare and suggest, but it also hallucinates a lot. I don’t think that’s a bad thing at all, precisely because it shows how important our role – as humans, but in our case explicitly as experts for innovation – remains.

In the next innovation briefing, I will take a closer look at this interaction between person and machine and the “human in the loop” principle. What remains our job, what do we delegate to AI – and how do we shape this new everyday reality responsibly?