The last few months have been busy. My colleagues and I in the LAB team (which is a bunch of AI engineers and innovators) have been in a number of meetings with people interested in our services. All the meetings have been enjoyable, as AI is something that innovation managers and business leaders love to talk about at length. However, when it comes to making a decision about investing in the research and application of AI in their company, many get scared.
It takes time to solve non-standard problems
Our starting point is always the same everywhere. We try to ask relevant questions and understand what the company or department is struggling with, what they want to improve. Then we work together to see if we can help them with AI. This step is often a revelation to the management of the company itself. One company sent us to their plant to talk to the people there. And it was brilliant, because we uncovered problems that didn't require AI at all. In another company, they had already identified the problem and suspected that it could be solved by applying LLM models. Great, something for us. A very specific problem. So we went straight to the initial validation of the problem. This is the step in which we check the current state of research and development on the problem and whether we are able to address the issue. Mostly in terms of capability, but also in terms of capacity and, of course, mental capacity. We also check if there is a much simpler or already finished solution to the problem (the so-called "off-the-shelf" solution) that we could recommend and do something more interesting instead. If we pass this pre-sales check, we prepare a proposal for the implementation of a proof-of-concept (PoC) solution.
Success is not guaranteed, but it can move you significantly
However, when we presented a Proof of Concept implementation proposal with the identified problem to the above company, we ran into a problem. The manager of the company expected us to give him some guarantees that the R&D we were proposing for them would be successful to some degree. Just that the Starship goes up and doesn't explode. But we couldn't give him that guarantee at that point. The nature of a research project is that it is not possible to guarantee its success in advance, or more precisely, the degree of success. Of course, we do not go blindly and, above all, we do not go into everything, but we go on the basis of our experience, what we know and what we can rely on. We left the meeting with the head down and the undesirable task of adjusting the price of the research to the risk of failure. He simply didn't want to pay the full amount if he wasn't sure of success.
Those who are really afraid, don't do R&D
As we later discussed the whole situation, we came to the conclusion that both we need to do a better job of managing the expectations of our customers (as well as our salespeople) about research projects, and that the path of R&D projects in new, unexplored areas is not for everyone. It takes leaders who are willing to go down the path of discovery, who are willing to take a reasonable risk of failure, knowing that if they succeed, they will not only have solved a problem, but they will have made their mark on the company as innovators. So it takes a bit of courage and, of course, a reliable partner who doesn't sell hot water, but relies on experience, a smart team, and knows how to guide the customer through new topics, such as the use of artificial intelligence in business processes.
A nice story we are currently experiencing with another international company in the industrial manufacturing sector, where we have found in the company director and his team exactly the discoverers who want to solve their long-standing problem with AI vision. They are not dreamers. They are pragmatic in their decision making, they have calculated the business case, they see the risks but also the opportunity to solve the problem and at the same time make their way through the group.
Our recommendations for managers:
- Try 1-2 smaller R&D projects to get started
- Choose problems for which you can calculate a business case
- If you are still worried, learn how to do R&D from more experienced people
Do you have an idea for an R&D project? Let's discuss it together!
