TwentyBN secures $2.5 million in funding for its artificial intelligence solution

0
281

The Berlin-based AI startup TwentyBN has just closed a funding round of $2.5 million to make the rapidly emerging AI method “deep learning” accessible to companies in all fields with a special focus on the European mainland and North America. The fresh capital comes from an US-based business angel who remains unnamed.

Deep learning systems are inspired by the neural networks in the human brain, and they have an innate ability to learn from data. According to TwentyBN, the method yields significantly better results than any other form of machine learning.

The aim of TwentyBN’s Central System Intelligence, under development since mid 2015, is to enable its customers to rapidly explore their applications using technical staff that are not necessarily expert in deep learning. The founders of TwentyBN want to enable customers across the world to create exciting new products based on deep learning after only a few weeks of initial integration work.  The TwentyBN AI lab is located in Toronto, Canada. Its research agenda ranges from advancing the art and science of network design with an emphasis on “recurrent networks,” to important technical issues relating to training deep learning networks, while aiming to advance the state of the art in the increasingly important areas of “transfer learning” where the objective is to learn multiple tasks at once.

Deep learning has emerged as the most effective approach to the creation of systems that solve real­world problems using artificial intelligence (AI). Deep learning is also key to many innovations, such as today’s early­stage self­driving cars, highly­scalable speech, image and video analysis, advanced text processing for language translation or to create dialogue systems such as those on mobile phones, and next­level medical diagnostics. On the horizon is also the very real possibility of effective robots to address many needs.

Chief Scientist and co-founder of the company, Prof. Dr. Roland Memisevic, explains: “Artificial intelligence is a discipline in which broad and deep experience is paramount. It is simply not optimal for every company to rediscover the same innovations in network design, training techniques, and build similar systems to handle immense datasets and computational tasks that are just as large. With our claim to deliver high­quality German engineering, we are making the latest AI techniques available through an extremely efficient technology infrastructure, while adding to these techniques ourselves in our research lab. We are making deep learning ready for large­scale implementation.”

Success with deep learning requires three ingredients: a compelling application, a solid scientific foundation with respect to network design, and a great deal of engineering to deal with the usually vast datasets, immense computations and complex industrial applications. Because of these stringent demands, and the need for high specialized scientific and engineering talent, the earliest successes have come from giant companies such as Google, Facebook, Apple, and IBM. Therefore, only a handful of companies have been able to profit fully from deep learning to date. TwentyBN wants to change that. The company is based in Berlin and aims to provide the scientific foundation and solid engineering to fuel the further growth of deep learning everywhere. The company is also exploring its own compelling applications.

The four founders, two of whom have resigned their professorships to devote their full attention to TwentyBN, met each other during studies at the University of Bielefeld in Germany. Each of the founders has over 15 years of experience in machine learning and the relatively young deep learning discipline. All of them have been programming since they were between 7 and 13 years old.

TwentyBN will operate from two offices, serving the European market from Berlin and serving the North American market from Toronto. With their first funding round concluded, TwentyBN now aims to rapidly expand its team in both research and technology.