Source: Deep Learning on Medium
How would you describe your business idea to a potential investor?
At Luminovo, our mission is to make artificial intelligence easily usable and widely used. We currently focus on B2B deep learning projects and building tooling to automate common deep learning workflows. As a start, we are rethinking repetitive workflows (like screening, monitoring, and labeling tasks) to get higher quality outcomes at a lower cost by having humans focus on the most difficult cases and letting continuously improving deep learning models do the rest.
What problem do you want to solve, what is your goal?
We strongly believe in the enormous potential of artificial intelligence to make everyone’s lives much easier. We want to work towards a world where humans can focus on the things they do best (things like creative problem-solving, moral decision making, and dealing with other humans) instead of on boring tasks, like sorting and screening thousands of images and documents (to pick a very specific example). Unfortunately, our world is messy and full of edge cases and the direct step from tedious human labor to full automation is often unfeasible. Instead of training a deep learning model with unrealistic expectations and being disappointed when it cannot deal with all possible inputs right from the start, we want to help companies take the best of both worlds from artificial and human intelligence.
How did you come up with your idea/concept?
We (Seba and Timon) met during our studies at Stanford University and while living together soon realized that we not only enjoy throwing large house parties together, but also both would stay up all night to nerd out about the newest machine learning paper. When coming back to Germany we decided to team up to help European businesses accelerate the adoption of deep learning. For the last 10 months, we have been slowly growing a kick-ass team and working on a diverse set of machine learning projects for clients ranging from small startups to big corporates. The idea of rethinking repetitive workflows with human-AI hybrids came directly out of our consulting projects where we saw that multiple clients were facing the same problems when trying to put our machine learning models into production. Rather than deploying our models in a one-off manner, we want to help our clients set up a sustainable process addressing the business problem they hired us to solve.
What is your business model?
For our projects, we charge daily rates or fixed prices like almost any other consulting firm, although every now and then we do performance-based pricing as well. On our product side, we still need to do more customer discovery and validate (or invalidate) our ideas before deciding on a business model.
Why did you start with consulting projects when you know it does not scale?
We really struggled with this question, in the beginning, especially coming from Silicon Valley where growth is everything. In AI the data you own is paramount, and data is really hard to come by as a startup. From what we have seen so far many of the best AI-powered business models create value without AI in the beginning and then become irresistible once they gathered enough data for the machine learning to kick in. It is a cheesy example, but Google could not have started as an AI-first company in 1998. After three months of exploration, we had talked to many companies that had valuable datasets and could really use our help to unlock their potential. We also really wanted to stop talking and start doing, so against the advice of many of the people we met we started running with the wolves and doing consulting projects last December.
It might be too early to tell, but we believe we made the right decision. We learned a lot about what it means to build a successful company (like hiring, management, and B2B sales processes), built some great relationships with our first clients, and learned how to build and deploy working deep learning models in the wild. And most importantly: we had a lot of fun doing so. We will keep you posted, but we think along the way we also discovered another path to a scalable AI-startup that does not rely on owning a unique dataset: building the tooling to let other people reap the rewards of their own data.
What is the best part about working at Luminovo?
The team! We are a pretty tight-knit group with a very active #random channel on Slack. We do regular company events together (just recently we went to Oktoberfest together, delivered a very average performance at one of the local pub quizzes, and stayed over-night at a winter hut during our hiking trip to the Alps) and our XPRENEURS office at WeWork comes with many perks like a barista, free tea, and even free beer. On a more serious note: We value psychological safety, honest feedback, and continuous learning (not just for our deep learning models) a lot and do regular 1:1s, retrospectives, and paper reading groups to reinforce our company culture. :-)
What is your top priority right now?
Our people are the most important asset to Luminovo! We are currently looking for the best deep learning engineers in Munich to join our team. If you know one of them you should definitely let us know. :-) We also have openings on the growth team every now and are secretly scouting for a technical genius to lead our product development efforts.
Why did you decide to work with XPRENEURS?
XPRENEURS was recommended to us by friends from startups from the earlier batches, and by people from the CDTM (Seba, one of our founders, is a CDTM Alumnus). Since XPRENEURS is all about building scalable tech start-ups with a repeatable business model, the program is the perfect fit to guide us on our journey from a boutique consultancy towards a product company. So far the exchange with the other teams in the batch has been very helpful, the WeWork office is super fun and the XPRENEURS team itself is always there to lend us an ear if we have any questions surrounding our efforts at Luminovo.
You want to become part of the XPRENEURS incubator program as well?
Get more information and apply at https://xpreneurs.io/