Listening Improves Artificial Intelligence!
Synergeticon: Hamburg-based startup relies on empathy for better IT
For a tech startup to be successful, it needs a very good idea, plenty of personal commitment and – once the orders start coming in – more and more experts who want to lend a hand.
Now, how do you go about recruiting AI experts, one of the most sought-after resources on the job market today?
“That’s indeed an enormous task” agrees David Küstner, who in 2015 founded Synergeticon together with Daniel Erdelmeier. “Nowadays, the really top-notch data scientists can choose from hundreds of job offers.”
Following the concept “experts want to work with experts”, Küstner and Erdelmeier first familiarized themselves deeply with the topic. As a result, even before they started their company, they were connected to various tech communities. They also needed this background to be able to identify the truly outstanding data scientists.
Once the project got underway, they actually succeeded in their first recruiting efforts. This again made it easier to find more staff of this excellence: "Our employees have a voice in the AI world, so they automatically serve their own networks", explains Küstner.
AI by Synergeticon: Interaction with the environment
Today, the startup is an established systems provider for industrial AI. IT solutions by Synergeticon support clients in digitizing processes by generating real-time models of production environments. A platform connects real world and model. The responsiveness of the connection is so fast that the user can interact with the environment safely.
A special forte of artificial intelligence by Synergeticon is object and event recognition in data generated by image-based systems like cameras and 3D sensors. There are plenty of potential uses: drone images improve maintenance at refineries, workbench cameras supply data for adaptive control of cobots, 3D sensors on workpieces permit in-line correction of machine settings, etc.
Retention Strategy: Personal development for technology freaks
Synergeticon wants to make digital transformation easier for its customers. For themselves, however, the AI experts are deliberately making things a little bit harder. This is part of their retention strategy.
Here Küstner and Erdelmeier act as decisively as they do in just about any other field. They want their staff to be able to live their AI and technology passion. That means landing orders that aren’t just work, but also lots of fun: “We only take on interesting projects, tasks that really challenge us. We all want Synergeticon to work like our own tech community, with access to the very latest in knowledge and research.”
Growing successfully: Effectively networked, but autonomous as a business
Evidently, the idea of the feel-good atmosphere worked out very well for the two AI thought leaders, because in 2019, just four years after Synergeticon was founded, the company achieved annual sales of 1.2 million euros. Only Corona could prevent it from becoming 2 million euros in 2020.
As one might suspect from their business address, ZAL TechCenter in Hamburg-Finkenwerder, one of the company’s first focal sectors was the aircraft industry. They’ve now branched out far beyond this market – without outside capital, as Küstner reports, and that’s how things are supposed to stay.
They prefer to keep relying on the bright ideas of their team, and on developing and maintaining partnerships. This approach has earned them a secure place in the aviation world – and an important companion on the journey was CENIT Hamburg.
CENIT as a startup partner
With over three decades of sector experience and our status as Tier 1 supplier of IT solutions for the major OEMs, CENIT was able to open doors for the young entrepreneurs. Küstner underlines that Synergeticon feels treated as an equal partner: “CENIT takes the long view, and you can feel it. We get valuable support, e.g. via concrete contacts to just the right specialists, or hints on how to best position ourselves within the corporate structure of the aircraft maker.”
In the same vein, the company was invited to join a key digital project. At CENIT this project is managed by Thomas Dulz. It was his idea to bring the AI experts from Synergeticon on board. “In aircraft manufacturing, things are constantly in flux. We are part of this process not just in concrete IT projects but also as a source of ideas. At CENIT, we therefore maintain close ties with applied research, but also with the startup world.”
The efficient path to certified AI
What makes Synergeticon a recommended AI partner for OEMs is the quality of their IT solutions, but beyond that the efficient way the startup generates robust algorithms for its clients.
For many businesses, projects like this are no longer some kind of futuristic experiment, but rather a way to exploit real ROI potentials in day-to-day business. To achieve this, AI has to attain standards that stand up under certification scrutiny.
And that’s also a question of how long it takes for artificial intelligence to be fully established, taught and tested. Küstner believes that Synergeticon has a technology edge in setting up training pipelines.
“In object recognition, you always encounter the same problem: for your training database, you need a slew of annotated data on the objects and events that matter to the client. Our advantage is that we are very good at automatically annotating such data, without a great deal of manual intervention. That makes us faster in setting up the data sets you need to create robust algorithms.”
Users aren’t in the way: They’re part of the solution
All the same, Küstner is convinced that each use case must be solved not just from the technology angle, but from a strategic one as well. And because Synergeticon’s systems often alter the working environment in their client businesses, dialogue with the client’s staff becomes a key aspect.
Consider a project in which cameras in a production facility monitor assembly progress at various workstations. This is highly sensitive information, because normally the cameras would also record the people working on the assembly line.
The technology solution is “ghost mode”: thanks to AI, anyone who enters the monitored facility is erased from the images – in the recordings, trolleys, components and tools seem to move on their own, as if by magic.
This solution was readily accepted by the client’s users because Synergeticon approaches project communication not as a formality, but rather as an integral system component. “We worked in close collaboration with the works council and the shop staff. Investing in dialogue pays off in a whole range of ways. The staffers feel onboarded and gain trust, and at the same time Synergeticon gets valuable information that helps us optimize our AI. This in turn gives the users a feeling of ownership, in the sense of ‘we helped develop this project and contributed our ideas!’”
Listening as a growth strategy
The idea that listening serves Synergeticon’s business interests is part of Küstner and Erdelmeier’s philosophy. “As a startup, we always wanted to develop solutions that the market needs. It wasn’t like we had a product and then went out to convince the customers that they should buy it.”
They still follow this approach today: “We concentrate heavily on client requirements and devote a great deal of time to fully understanding the problem and how to solve it.”
This sort of mindfulness is good strategy these days, and it’s a formula for success that can help other startup enterprises as well.
What other hints can they offer? “Seize every opportunity to take a break!”, says Küstner. “Running a business is very demanding, and those demands can start to dictate your whole life. That makes it all the more important to make the most of your downtime.”
REGISTER HERE FOR THE CENIT MAGAZINE
Subscribe to the best technical articles and blog posts from the CENIT Newsroom plus additional updates from our company.
The CENIT Magazine brings you fresh thinking and clear messages to help you successfully manage digital transformation.
Don't miss an issue now!