| Biotech | Chemistry | |
|---|---|---|
| Access to capital | High; we see high levels of funding & competition in the space across the foundation model, application, and full-stack R&D vendor layer. Mostly due to the category being much more mature and therefore well ‘understood’ by investors. There is a more established universe of specialized funds focusing on these opportunities. | Medium; the end market is still much more nascent, and so far we have not come across many new competitors with significant funding other than Orbital Materials. Due to that, the broader investor universe is hence much less developed at the moment compared to the biotech/biology world. However, this can also be an advantage, as there might be more excitement to back one of the early, pioneering companies in an emerging sector. Besides generalist funds, this sector should appeal to the broad base of deep tech funds across US/EU. |
| Buyer Readiness | Buyers are typically more comfortable with leveraging generative AI due to their increasing familiarity with traditional AI driven drug discovery techniques. There is industry-wide recognition that transformer models and others can play a significant role in accelerating drug discovery and other R&D efforts. Some early proof / ROI has been established around candidates discovered or refined with the help of AI. |
However, readiness to engage directly with foundation model providers remains mixed. Some pharma companies and biotechs have robust data & AI teams to run and fine tune models, but most prefer a more high touch engagement with model or solution vendors. | Buyers are aware of the potential of transformer models. However, many experts we spoke to expressed skepticism that materials scientists and other key R&D stakeholders are receptive to new technology. The cultural hesitancy about new technology and R&D techniques seems to be a meaningful challenge in this market.
Most experts had a preference for a partnership / co-development model and recommended focusing on quick time-to-value use cases to gain an initial foothold with large enterprises.
Still TBD weather there will be a ‘biotech’ moment for chemistry or weather the GTM will remain different due to a slightly less forward-leaning buyer persona. | | Data Availability | Compared to material sciences, data is more readily available and structured. This is particularly true with protein folding data. In other modalities such as histology data, data remains siloed across research organizations, pharmaceutical companies, academia etc. However, data is typically already digitized and the unlock is moreso around building the right partnerships to acquire data across organizations and modalities. | Significantly less mature compared to biotech & pharma. Many experts we spoke to indicated that data from previous experiments is often unstructured and at time still on pen & paper. Many large materials businesses like Dow are still in the midst of digitizing their data. Furthermore, R&D experiments in this sector are comparatively less structured, making high quality longitudinal data harder to find. | | Potential Time to Value | Long due to stringent regulatory scrutiny and multi-year clinical trial timelines, however, clear and well understood gating moments along the way, that lead to significant value increases when they are passed. Prize at the end of the line tends to be very large. | Potentially much shorter, particularly for developing incremental improvements on existing materials. Though compliance and regulatory approval play a role here as well, they are relatively lighter compared to biotech, but value of individual molecules tends to be more varied |