“I had a corporate job that was exciting in terms of content, paid accordingly and not that easy to leave. I needed a big challenge to escape from the golden cage,” says Martin Nigsch, who founded the Vorarlberg-based deep-tech startup feld.ai in June 2022.
The decision to start a company was not an easy one, even if the Vorarlberg native was itching to do so: “You are seen as a nutcase if you simply give up a job that others can only dream of as a career goal,” says Nigsch, who previously worked full-time for a large corporate insurer.
The father of three children did it anyway and founded his startup feld.ai. He began to explore the domestic startup and funding landscape in autumn 2021.
When he Preseed-DeepTechprogram from Austria Wirtschaftsservice (aws), he dared to take the first step out of the golden cage mentioned above. Very soon after the aws funding, feld.ai received its first orders, which have sustained the company to date. The core of the business: an “operating system” to solve difficult challenges with documents in companies.
Money was only “one factor”
There were several reasons for the launch of feld.ai. One of them was the financial support from AWS. “The funding from AWS was a great springboard,” says Nigsch about the beginnings. However, money was not the decisive factor that made the launch possible. The first priority was the idea and the awareness of the size of the problem in a rapidly growing market.
“The AWS gave me the stamp: ‘He’s not completely crazy’”
“The first thing you need as a founder is confirmation that the idea is good and that you are the right person to implement it. That’s what I got from AWS at the beginning: The decision-making body expresses this trust, AWS helps you at the beginning, sets milestones together, monitors them, then withdraws at the right time and otherwise does not interfere operationally. In that sense, I got everything I needed in this phase,” says Nigsch.
Pat on the back for data management
What AWS helped with to the right extent was the development of an AI-based document management system for companies.
Nigsch had already recognized the need for such a service at his previous employer: “Feld.ai emerged from a corporate context. I previously worked for a corporate insurer where data is the necessary basis for concluding profitable business.”
“The problem we started with was extracting structured information from documents. But mere extraction and storage is not enough. You always have to put information into context and analyze it. This is especially true when companies need to use documents as a basis for process optimization and decision-making,” says Nigsch.
This is where feld.ai’s solution comes in: The startup helps companies from various industries to unlock the “gold treasure” hidden in their documents.
More specifically: “We help companies become more efficient and make better decisions with the help of AI. We focus on document-based company processes.” These include: emails, customer inquiries, invoices in the form of PDFs, audio files, Powerpoint, Excel, Word, text and images.
“The inbox is an unsolved problem for many customers”
Feld.ai covers several application areas. One of them is easy to explain: the management of email inboxes: “Surprisingly, automating and classifying the inbox is an unsolved problem for many customers,” Nigsch reveals from experience.
The interesting thing about it: Feld.ai not only stores data in a structured way, it connects it to existing information, analyzes it and draws conclusions: “Basically, the things we do can be solved manually and without AI on a small scale. But if you have classified, tagged, automated and linked an unmanageable amount of emails and customer inquiries with inventory data, analyses, process optimizations and thus efficiency in the company are much better.”
Incoming mail and cost accounting
Practical applications lie not only in incoming mail, but also, for example, in the automated posting of cost invoices: Individual invoices or customs documents are read out using AI, linked to inventory data and processed accordingly.
However, the exact work of feld.ai is not easy to explain. After all, the company is dedicated to customer-specific challenges, including handwriting, inconsistent nomenclature and different languages. The scalable core is that the technical problems to be solved are very often repeated, regardless of the industry: The challenges always revolve around the structure, analysis and keywording of documents and how the information obtained can then be integrated into further processes.
Data backup without American mothers
Now you might think that there are already plenty of data managers and analysts on the market – especially on the North American market. That may be true, but it often raises questions when it comes to data protection and data transfer. It is not always clear what happens to data stored in a US provider’s cloud.
Here, feld.ai relies on a regional solution – far away from the giants OpenAI and Microsoft: “Data security is a big issue. We offer our own solution for this, in which data is backed up in our data center in Vorarlberg or locally at the customer’s location. There’s no way we can go to an American parent company.”
No OpenAI and Microsoft, but a Feldkircher Box
When it comes to using AI, the company relies on local, “smaller, more efficient language models”, trains them and develops them. OpenAI, Google, Amazon Web Services and Microsoft are left out. To ensure data security, feld.ai operates its own data center in Feldkirch, Vorarlberg. Data security “in our box” is therefore guaranteed.
But this is not a black box, because feld.ai relies on transparency: “Since day one, we have been building on producing visible efficiency results.”
Meanwhile, feld.ai offers a three-part product portfolio: Firstly, as an “as a service” product in the sense of a company solution managed entirely by feld.ai. Secondly, as a “managed service”, which is installed on customer infrastructure and managed by feld.ai. And finally: “managed by customers”. The feld.ai solution is installed on an existing IT infrastructure – locally or in a cloud – and finally also operated by the customer’s AI or infrastructure teams.
Already listed partners
They are now the official technology partner of the Swiss company Kendox AG – an expert in software and data management. They operate in the B2B2B area. The partner company supports 1,500 corporate customers with over 600 million documents. “For us, that means: the ideal first partner,” says Nigsch about the partnership.
However, the Feldkirch-based company is not focusing on a core industry: the startup already has customers from the auditing, insurance and legal sectors. Promising discussions are currently taking place, especially with German and Swiss customers: Austria is (still) underrepresented.
Feld.ai charges a “small fee” per document. Growth and scaling are planned and necessary because “the whole thing won’t work if it stays this small.” The company currently has three full-time positions and four part-time employees or interns.
Opportunistic and organic
The fact that Nigsch, a father of three, dared to leave the golden cage and try his luck with his startup dream is thanks in particular to AWS. The support from family and friends also contributes daily to Nigsch living out his dream of having his own startup.
In the future, feld.ai wants to continue to “move opportunistically” and grow organically in order to continue to grow. “We try to stay on the golden path that allows us to continue to develop while creating value for our customers.”
Link tipTo the startup’s website
*Disclaimer: The startup portrait was created as part of a media cooperation with Austria Wirtschaftservice (aws).
