Investing In Reliant AI: Empowering Knowledge Workers In Biopharma

Nearly two years since the initial “LLM moment”, we continue to see an explosion in both startup activity and enterprise adoption. 2023 accounted for over $21B invested into AI startups – $15B from strategics, $6B from VCs across 721 deals – representing a 5x increase from 2022. The proportion of S&P 500 companies that mentioned AI on earnings calls hit an all-time high in Q4 2023 at 36%. 

It is an extraordinary time to invest in innovation and disruption. 

At the application layer, we have invested across two archetypes: experimentalists and technologists. Experimentalists (like our friends at Spellbook) ride the wave of generative AI technology by relying on third-party models, prompt engineering, workflow integration, and information retrieval. Conversely, technologists customize open-source models or build models from scratch to own the inroads to customer workflows. As discussed in our whitepaper, these two approaches come with different risk profiles spanning barriers to entry, capital intensity, and defensibility.

For us at Inovia, the technologist approach at the application layer is generally an exception. In order to invest, we must see unique insight suggesting a reason to own the model layer (versus riding a technology wave), and we must see a world-class team at the helm, armed with the ability to build truly differentiated technology. Reliant is such a team.

Today, we are excited to announce the latest technologist addition to Inovia’s AI portfolio by leading Reliant AI’s $11M Seed round alongside Tola Capital. Reliant, a genAI-powered data analytics software provider that accelerates secondary research, is building the AI workflow platform for biopharma, starting with their first product, Reliant Tabular.

Reliant co-founders are leaders in reinforcement learning. Karl Moritz previously exited his NLP startup to Google and co-founded DeepMind’s language group, while Marc published industry-defining work in the Journal of Artificial Intelligence Research, Nature, and MIT Press and led reinforcement learning teams at Google and DeepMind. Karl Moritz and Marc have the technologist toolkit we are looking for to build a world-class product, bring on board exceptional talent, and tackle a complex problem.

We believe that commercial pharma warrants a technologist approach, given the prolonged technical differentiation required to maintain long-term value. Pharma workflows today are menial and repetitive, facing headwinds from the rising cost and time to commercialization of pharmaceuticals, and the proliferation of unstructured data. Despite their inefficiency, these workflows are highly specialized and require routine aggregation of technical information from disparate, industry-specific sources. Although off-the-shelf commercial models tout benchmark performance ranging from 75% to 90%, that’s  not enough to account for biopharma-specific nuances. Reliant is developing a proprietary machine learning system that is optimized for high-accuracy results to free knowledge workers from the mental menial labour that takes up their entire day, Reliant’s system already achieves close to 99% accuracy for select biopharma-related workflows – a phenomenal improvement over the baseline. As Reliant expands across commercial biopharma, the impact on workflows will resonate from medical affairs to market access, ultimately driving critical ROI across an enterprise. 

We are doubling down on Reliant, and can’t wait to continue working alongside Karl Moritz, Marc and the whole team as they build the next generation of tools to innovate and disrupt the world of commercial pharma.

---