Leveraging AI to bring B2B transactions to the digital age — Why we invested in Workist
We are excited to have led the €9 million Series A financing round of Berlin-based Workist, joining existing investors including our friends at LEA Partners, 468 Capital, and another.vc. The stellar team around the founders Tim Wegner, Alexander Müller, and Fabian Brosig built an AI startup automating document-heavy B2B transactions by offering a super-smart translation layer between the company’s different ERP systems that are structurally disconnected.
The landscape for Enterprise Resource Planning (ERPs) tools is extremely fragmented and firms’ tech stacks notoriously disconnected. To put it simply, every business speaks its own language when it comes to IT and data. Since the birth of electronic data interchanges (EDI) in the 1970s, there was hope for a standardized way to exchange business data between two firms and with that, a foundation for automated transactions. Yet, more than 50 years later, EDI adoption is still low with little perspective that this will drastically change in the future.
🔄 💾 Business transactions are stuck in the pre-digital age
Today, most business transactions (incl. orders, invoices, and procurement-related documents) in companies (from small to enterprise) are still highly manual. Once transaction documents are received via email, pdf, or xls, people must manually open the documents, read through them to find and copy the relevant information (e.g. item number, order quantity, price, date, customer id), and then, physically type or paste them into an ERP system. This manual process is a time-consuming, often frustrating job, as well as inefficient and error-prone.
🙇♀️💔 Existing solutions fail to address SMB needs and capabilities
To escape this unfit internal process, a company currently has 2 options:
1️⃣ Outsourcing: this might make the process marginally less expensive, but it is still not a scalable solution. More importantly, by not having your own team extract the unstructured data, your firm loses the ‘implicit knowledge’ they have about your products and your customers: (e.g. customer X has a special customer ID, and when they order what they internally call order item #10, they mean our order item S43J.)
2️⃣ EDI: in theory, it’s a great way to streamline data exchanges. The problem here is that companies need to manually build point-to-point integrations — quite a painful task — to all parties in their supplier network. Even for SMBs, this can easily amount to several hundreds or thousands of business partners. This explains a very low EDI adoption where companies usually just build direct EDI connections to a selected handful of business partners with the highest transaction volume.
💡🌊 A first wave of start-ups tackled the problem, but only for a selected few
Driven by the emergence of optical character recognition (OCR) allowing data extraction from unstructured documents, as well as robotic process automation (RPA), a first wave of companies began tackling the problem of B2B transaction processing.
On a technical level they work and do the job for basic, static use cases. This is why some of them are now well capitalized and well established on the market.
On a more pragmatic level however, they failed to offer their tech as a sophisticated product that companies are able to integrate and use on their own and are thus dependent on integration partners and consultancies — a long and expensive way for customers to get started with automatic document processing. This already is a hard pill to swallow for enterprise clients with the budgets and capacity for these sorts of large transformation projects, but it is definitely an adoption killer for SMBs.
They also come with another huge drawback: they are missing the above-mentioned implicit knowledge a human worker has when moving through orders of a client they know well, for products they know well.
🔀 🚫 A parallel trend — the anti hypothesis:
In recent weeks, there has also been a lot of buzz around a different approach to solving the same issue: a platform approach where an intermediary builds and collects all direct EDI intersections to big suppliers and makes them accessible to others.
Even though we at Earlybird Digital West believe in an increasing standardization of data exchange in many different industries and multiple ways, we do not see it as feasible for this use case.
Firstly, supplier connections are mission-critical so companies are not well advised in giving away ownership of these to an outside player. Additionally, this approach works best for narrow markets with a few selected players on top, into which everyone has to integrate (e.g. food retailing in Germany), but not for more fragmented markets with a large mid- and long-tail. We also believe that due to the fragmentation of the ERP landscape, the uniqueness of business processes and the idiosyncrasies of market participants, there will always be a mismatch between internal and external classifications, for which a super smart translation layer in between is needed, instead of a static data exchange pipeline.
🚀 🎬 Clear the stage for Workist
Workist is a next-gen verticalized RPA solution that not only mastered the data extraction from unstructured documents, the validation, and ingestion as part of B2B transaction automation, but also managed to put it into a sleek and user-friendly product. It is a true plug & play solution with quick time-to-value, skips integration work on the customer’s end, and requires no developer capacity nor special skills. It achieves high automation rates from day one that continuously increase due to the application’s self-learning AI algorithm– observing and learning with a human-like implicit knowledge component. This comprehension of the system can then be used to even better validate the extracted data, as well as correctly ‘translate’ mismatched data entries or fill missing gaps.
We are proud to be leading this round and join the team on its mission of bringing B2B transactions into the digital age.
Workist is looking for talent in roles ranging from engineering to sales — you can find open positions here if you want to join this incredible team on their journey.
Team Workist, welcome to the #EBVCgang. We’re excited about the journey ahead! 🚀🚀
By @LaurinClass and Andre Retterath
