From the collection of trash to visual manufacturing inspection and dynamic pricing: Artificial intelligence offers potential solutions to many challenges in a variety of areas and therefore opens up many opportunities for the startup community. 

When it comes to building AI-startups, founders find a specially nurturing ecosystem in Bavaria, Austria and Switzerland: access to talent through universities such as the TU Munich or the university of St. Gallen. And a strong tech ecosystem due to many of the major tech giants having huge offices nearby, including Google, IBM and Microsoft as well as the big corporates such as BMW, Siemens, Airbus or Allianz – a very good breeding ground for the next generation of AI leaders. 

That’s why we’ve decided to regularly present the league of the most promising AI-newcomers in the startup ecosystem from Munich, Bavaria, to Zurich, Switzerland. Let’s have a look!  

Angsa Robotics *

Angsa, which is getting funded these days, is developing a robot that automates garbage removal on surfaces such as grass or gravel. Conventional sweepers can’t be used there so Angsa's core innovation is the development of an artificial neural network that can locate even small pieces of waste such as cigarettes or bottle caps in the grass with a camera. 

The autonomous robot then removes the recognized objects without damaging the surface or destroying insects. The AI also distinguishes between drivable ground and obstacles according to Karl Schulz, who is one of the founders. 

Business model: Angsa’s entry-level market is the use of the robot after festivals or other major events, where cleaning must be carried out thoroughly and under great time pressure. Customers are cleaning service providers, to which Angsa rents several of its robots to carry out an order. The second application is the permanent use in parks or outdoor pools, where the sale of individual robots is the most suitable for customers. The long-term plan is that customers pay a fixed amount per day if they rent the robot. 

Amount of funding: Angsa Robotis is currently funded by a start-up grant. A pre-seed financing round of over 100.000 euros is planned for autumn 2020.

Number of employees: The core team consists of the three founders and they get support from eight TUM students, who help develop various robot modules.

‍

FotoNow 

Fotonow is a Munich-based startup which was founded in August 2018 as a consultancy with a focus on AI and computer vision. Since October 2019 the company pivoted from a consulting business to building a product: The company now focuses on an AI-based visual manufacturing inspection, which detects errors in the entire field of vision of any camera, regardless of the location. 

“This brings quality control directly to the source of the error”, says co-founder Florian Ziesche. In order to do that, the team has trained more than 1.000 AI-models. 

And the startup has high ambitions: “We want to expand our team with the best and brightest minds in the world - and to become, like Celonis, an internationally successful unicorn from Munich”, Florian Ziesche stresses. ‍

Business model: FotoNow offers a software that is sold as a license which is paid annually.‍

Amount of funding: The team received 75.000 Euros as part of an accelerator program and wants to open the first major funding round in October 2020.‍

Number of employees: Apart from the two founders the team consists of 1 intern, 5 part-time employees in the fields of marketing, business development and technology, several coaches, advisors and a business angel.

‍

NeuroForge 

NeuroForge is an AI-startup based in Bayreuth, which was founded in January 2019. The company develops artificial intelligence for manufacturing companies with a focus on quality assurance and predictive maintenance. At the moment the startup focuses especially on small- and medium size companies in their region. 

One of its products, NeuroForge Compose, is a platform for centralizing and uniting process and company data. Aside from building a company it’s NeuroForge’s bigger vision to build more trust towards AI technology. ‍

Business model: NeuroForge’s business model consists of two pillars. The first one is advising their customers in the manufacturing industry on how to use their AI-applications. And secondly, they sell their data warehouse product as an IT service.‍

Amount of funding: So far the team bootstrapped NeuroForge.

‍Number of employees: 6 people 

‍

presize 

Presize is a Munich-based startup founded in April 2019, whose technology generates a 3D-model for e-commerce customers. The technology creates a model of a person from a short smartphone video and uses this model to determine body measurements.

These body measurements are then converted into a size recommendation for fashion online shops - based on product data and transaction data. Presize uses AI in several ways: The algorithm has learned what human bodies look like and can therefore develop a 3D-model. 

Using this 3D-model, the team can determine body measurements based on a very large data set and its AI learns from the transaction data (return yes / no) how good its size recommendation was. The team's goal is to enable online shoppers one billion perfectly fitting product purchases by 2023. ‍

Business model: Presize targets to increase the conversion rate of online-shops and to reduce return rates. They earn a commission if a user orders based on a size recommendation.‍

Amount of funding: Presize is currently completing its second round of funding over 1 million Euros. Their investors include Plug & Play (PayPal, Dropbox, N26), Christina Rosenberg (ex-CEO Hermès Paris GmbH), Chris Brenninkmeyer (part of the C&A family)‍

Number of employees: The team has grown to 15 “presizer”, as they call themselves

pricenow 

Pricenow is a Swiss startup, which was founded in 2017 and focuses on price optimization for the tourism and travel industry.

The team uses AI for its main product, called Dynamic Pricing, and especially machine learning models, to make the most accurate demand forecast possible. 

“For our customers we update 150.000 price points every day”, says co-founder Judith Noerpel-Schneider. In addition, Pricenow also uses machine learning models to determine the different factors of price-demand elasticity. The team is currently working on its market entry into other industries. ‍

Business model: The business model is based on a license. The startup offers different services: dynamic pricing, data analytics and the programming of e-commerce websites. The team also works as a consulting company on a project basis.‍

Origin: Reichenbach/Bern ‍

Amount of funding: Pricenow has completed a second round of funding, but doesn’t disclose the amount.  ‍

Number of employees: 9 team members 

sinpex 

Sinpex is a Swiss-German startup, founded in 2019, which created  self-learning algorithms for the extraction of data from ressources, for example withdrawing knowledge from unstructured and structured text documents.

“We use a combination of natural language processing and deep learning, which is able to evaluate any type of document”, says co-founder Camillo Werdich. That’s unlike other bots, that are only applicable to certain document types or databases, according to him. For example Sinpex can answer customer questions in standardized online-questionnaires in the banking area, and then document the answer comprehensively. ‍

Business model: Sinpex can be obtained as a monthly or annual license and the price depends on the number of documents processed and the number of questions to be answered. Sinpex can be obtained directly from the startup or in larger projects through its partners, Horváth & Partners or PWC Germany. ‍

Amount of funding: Sinpex received a total funding amount of 500.000 Euros. ‍

Number of employees: The team consists of three members and is supported by freelancers.  

‍

*We’ve decided to present the startups in alphabetical order.

‍