Some Weeks are just...

Busier than others. It looks and feels like we are metaphorically on fire. Or from an outsiders perspective they appear to be. This week was one of those weeks, for 3 pre-seed stage investments we at Twin Path Ventures led on, over the past 12 months, all got announced. HotHouse Therapeutics, Azoma and Mater-AI all came out of stealth this week to announce their rounds ( ranging from $2m to $4m) and to reveal their plans. As, you will expect for deals from an investor in AI first/ AI Native startups, all three of these portfolio startups are co-led by world class scientists/data scientists and are all on a mission to become category leaders in their space. And all are doing this, in their own way, by purposely building "State of the Art" AI platform solutions able to generate novel assets that are the most capable on the market. The fact that all three startups look very different shows in a small way how diverse the use cases are for the latest generation of AI is and the opportunities it can generate for investors. Below is a short rationale for our investments in these diverse startups and the reason behind what knits them all together within our investment thesis.

Hot House Therapeutics is primarily a bio-tech company developing sustainable, scalable pharmaceutical compounds through AI-optimised plant-based chemistry. Their initial focus is deploying their AI platform to develop an alternative and better performing adjuvant used in vaccines to boost the immune response of patients and therefore the vaccines effectiveness. At present there is a global dependence by vaccine makers on a single, unsustainable source of the critical adjuvant, QS-21, which is a natural ingredient extracted from the bark of a slow-growing Chilean tree. The world needs an alternative, especially fingers crossed, the next generation of vaccines for cancer treatment hit the market in the next few years.

The HotHouse Therapeutics AI platform can currently search for and unlock over 20 million novel molecules by combining AI-driven pathway prediction with the enzymatic diversity of the plant kingdom. Once they find a novel molecule they bio-engineer ( i.e. grow in plants) the compound in their Norwich UK facility, purify it into pharmaceutical grade material and test its medicinal properties in CRO labs. If that all goes well ( and so far-so good), they intend to sell their patent protected novel adjuvant ( or a licence to re-produce) to the vaccine makers.

While the initial focus is on vaccine adjuvants, the company is already planning its first traditional drug discovery programmes, using its AI platform to identify and then progress assets through to clinical development. So the AI platform makes Hothouse more than a biotech startup selling small number of assets. Hothouse Therapeutics have a "repeatable" capability to syntheses SOTA AI with bio-engineering. This is why Twin Path Ventures (primarily via Katie Lockwood ) led the round and why we believe these new hybrid AI-Biotech startups are so game changing for drug discovery. They can do it again and again.

Mater AI ( https://mater-ai.com) is also building a discovery platform using the latest SOTA AI models and thinking but in their case it is to find and then synthesis in lab, new advance materials with special thermo electrical properties. Materials that can turn heat into electricity or provide cooling without refrigerants. Their AI and physics-based modelling platform is designed to accelerate the finding, design and lab verification of novel synthesised new materials, from decades to weeks – up to 438× faster than traditional R&D – optimising at the design stage for properties like thermal and electrical conductivity to achieve higher efficiency, lower cost and ease of production. Another great example of how a team of PhD's from their respective scientific fields can come together to use this generation of AI to build the next generation of technology. We at Twin Path Ventures invested because again we can see how their in-house AI expertise gives them a massive advantage over competitors in the space. Giving the once in a generation investment going into building energy inefficient data centres, never has there been a better time for new thermo electric material able to "make heat useful again"

Finally Azoma, which on the face of it you may believe, has little to do with the scientific discovery engine startups described above but you would be wrong. Firstly the team is highly technical, with the CTO Timur Luguev leading a team of data scientists and engineers in Toronto. This team has worked together for almost 3 years developing a now patent protected Loss-Guided Diffusion Model that alongside their fine-tuned LLM and auto-prompt embeddings ( and a great deal of data and industry knowledge) has enabled them to build a SOTA Generative Optimisation Engine that actually delivers for brands and e-commerce sites. In a sector, where the change from Google search to LLM powered research has incentivised a lots of "thin wrap" startups to emerge promising clients that through simple prompt engineering they will reverse the decline in traffic from standard SEO, Azoma stands out. They stand out because they have thought deeply about the issue, put in the hard miles to build a bespoke AI solution to the challenge, have assembled and put together a world class AI team that will keep on innovating, adapting the best in AI R&D and as such they have a lot in common with the startups that Twin Path Invest in.

We at Twin Path back AI first startups with the capability, the ambition and hard graft to produce solutions that are novel, innovative and potentially the best in class solutions to big challenges faced by customers they truly understand. We invest before these solutions are validated. For the likes of Mater AI and Hothouse Therapeutics the initial validation will be in the lab, proving the efficacy of the novel assets they have discovered and produced against known scientific benchmarks. Mater AI are early in this journey but moving super fast, whilst Hothouse are now close to being able to show the world some very impressive results.

For Azoma the validation is more straightforward- will brands and e-commerce sites use Azoma and will they see value. For Azoma the results over the past 12 months are truly impressive with retailers such as Canadian Tyre and brands such as Mars and P&G all confirming that Azoma solution is best in class.

So all three have come out of stealth this week because they have been building and validating their foundational approach to building best in class AI platforms. Yes it takes time, specialist expertise and a lot of hard work ( and a burning passion to succeed). For that they do deserve a round of applause even though their journeys to become category leaders has only just begun.

So yes this week is a busy week for Twin Path Ventures, but not as busy as it for our portfolio company founders. They are on fire.

Finally, Twin Path Ventures has now invested in 32 AI first startups since July 2023 and we are actively looking for more. Reflecting on this experience, means that I am in the process of compiling a report/ presentation for our LP's and investment partners on what we now think is going to happen in the next few years with AI and our reasoning and evidence why some categories of AI first startups are well positioned to win. Happy to share on request with prospective partners and investors.