Overheard in Maribor, Slovenia
Last week Twin Path was a proud attendee of the PODIM tech conference in Maribor, Slovenia. Despite being an investor in predominantly UK startups, we were delighted to participate, give pitch feedback and get a fresh perspective on the ever-evolving tech landscape. With huge thanks to the organisers, we highly recommend this to anyone in the tech investor ecosystem with a penchant for quality over quantity. We found ourselves returning with our hands brimming full with meaningful connections and a good grasp of what AI founders in CEE / Balkans are building. Here is a run down of what we overheard in PODIM.
AI startups & the purpose of revenue
At a lively panel event, investors discussed whether startups, even AI startups, need to focus on obtaining revenues and early signs of profits before being investment ready. Some of the panelists vehemently argued that repeatable revenue is by far the best signal that a startup can reach product-market fit and thereby scale and raise a good series A and beyond. Indeed, with the AI space being so hot and hyped right now (valuations over 2.5x for AI than B2B SaaS at Series A) we can understand the appeal of using revenue as the ultimate yardstick of value.
The view of Twin Path, as espoused by our partner John Spindler and neatly summarised by Romeo Walters of Fly VC, is more nuanced. The purpose of all companies must be revenue generation (not of course to the exclusion of other goals). However, the revenue profile of different companies can vary hugely. For deep tech startups including Frontier AI, more important than early revenues is that the startup can build and prove in the field that their innovative and novel technology works, solves a big non-trivial problem for their customers and creates and captures massive value. In regulated markets, especially health, the effort and time to validate the efficacy and safety of the innovative solution can take years. The most important metric for investors is that the problem is real, can't be adequately solved by existing technology and that there are at least key opinion leaders from their target market helping the startup with offers of data and design partnerships.
For more applied AI solutions, the race is on to leverage AI to solve known problems, for known customers and to become the market leader. Here, early traction is vital and by far the best evidence of the startups probability of achieving PMF. The product development period prior to launch and early traction is months not years. Nevertheless even in these cases, not all customers and revenue are equal. Less income but from the right customers who validate that the AI powered product can and does create enormous value for the target client is preferred. Conversely, a few lucrative contracts of the wrong type can end up being a distraction. As this diagram illustrates, we at Twin Path want to see that a startups’ AI solution (which will be more costly to build, test and deploy that rules-based software) can be disruptive (and that that disruptive competitive advantage can be preserved).
The trick then is to pick the best signals for the stage and type for the company and in certain cases there are better signals than the amount of revenue you are generating.
Interesting AI startups that we met
Katie Lockwood (Twin Path Partner) on the judging panel at the PODIM pitch finale
Ulpian AI - super smart legal tech startup combining expertise in law, AI and software engineering to seamlessly integrate technology with legal practices starting with civil law jurisdictions.
Panda Chat AI - this started as a hobby project of CEO, Primož Cigoj, over the weekend in Ljubljana, Slovenia with the aim of making chatting with data easier than ever.
Novasign - A bootstrapped startup out of Austria that is combining the power of AI with Mechanistic Knowledge to predict, optimise and control both bioprocess development and biomanufacturing in a better and faster way. They have an impressive number of clients and are now looking to raise a late seed/ early Series A to further build out their platform and commence scaling.
Battery Check - We are seeing great opportunity in the "2nd Hand/ pre-loved" industrial battery sector whether those batteries are in Electric Vehicles or powering off-grid deficies. The major barrier to the development of a 2nd hand battery industry is analysing and predicting the likely future performance of an industrial battery, its charge, discharge and anomalies now and in the immediate future. The present tech for testing is crude and manual. It seems like a perfect use case for AI. Battery Check, is a Czech-based startup offering such a solution in the region with ambitions and plans to go worldwide after they raise their seed round. It will soon be a very competitive space but this team has a good combination of PhD AI expertise and commercial experience to succeed.
