Zephyr AI Raises $18.5M Seed to Advance Precision Medicine using ML
On average, it takes 10–15 years to discover & develop a new drug. But what if we could do it in a matter of days?
On our latest episode of The Breakouts Podcast, we asked David Morgan and Jeff Sherman — CEO and CTO who just raised an $18.5M seed round for their AI startup, Zephyr AI.
But First, Some Context:
Zephyr AI is building data & analytics tools to address unmet needs across the full spectrum of precision medicine — from prevention to treatment.
What is precision medicine?
It’s a type of medicine in which medical decisions, interventions, & treatments are personalized to individual patients based on their predicted response to treatments.
Zephyr AI has 2 key product pillars:
Treatment:
- Drug discovery
- Drug repositioning
- Clinical trial design
Prevention:
- Population health management
- Personalized patient health
- Decision support
On the treatment side, it builds internal research tools that can predict a patient or molecule’s pharmacological response.
It co-develops drugs with other companies and uses sensitivities outputted by its tools to accelerate the process.
On the prevention side, it builds a predictive analytics SaaS tool for healthcare institutions so that they can offer personalized interventions, risk assessment, & decision support to patients.
So what can we take away from Zephyr AI’s approach to advancing precision medicine?
Onto our top 3 insights 😎👇
Insight #1
Advances in ML could make drug discovery a tractable problem.
Making quality predictions for drug discovery requires analyzing data across a large number of dimensions.
“It’s beyond the comprehension of the human mind to be able to look at all the factors in aggregate.”
But it may no longer be beyond the comprehension of computers.
Over the past 10 years, we’ve seen major leaps in:
i) Compute power
ii) Compute cost
iii) Sophistication of AI/ML models
In aggregate, these headwinds are beginning to make it possible to solve problems in drug discovery & medicine that we previously thought were intractable.
That’s Zephyr AI’s “why now”. And it could be yours too 😉
Insight #2
“Failing as cheaply as possible is just as valuable as success.”
The average cost to develop a new drug and get it to market is $2.6 billion, and the average time taken to do so is 10–15 years.
Every incorrect iteration during drug development is immensely costly.
Zephyr AI:
- Aggregates data from data partners
- Feeds data into ML models
- Uses ML models to output sensitivities
- Uses sensitivities to determine a more intelligent next drug iteration to test
- Slashes drug development costs (due to intelligent iteration)
Insight #3
When building in #DeepTech, find clever ways to be lean.
90% of drugs fail in clinical trials and never make it to market. In #biotech & #pharma, the probability of failure is high and so is the cost.
To beat the odds, you need to reduce the time & cost taken per iteration.
Zephyr AI is prioritizing use cases that require lower time per iteration.
Ex: rather than starting with creating new drugs alone, they’re first using their technology to find new applications for existing drugs — so it doesn’t take 10–15 years to realize the impact.
And they’re also finding ways to reduce the cost per iteration.
Rather than developing new drugs alone, they’re co-creating drugs with other companies.
These companies pay them milestone-based service fees as well as a residual licensing fee once the drug is on market.
The milestone-based fees offset the cost per iteration.
And since Zephyr AI partners with other companies to create drugs, they have enough bandwidth to co-develop multiple drugs in parallel.
Each drug that makes it to market makes them money through residual licensing fees.
Similar to a VC portfolio, they only need a couple of their bets to pan out to justify their investment.
And thus their partnership model mitigates the traditional 90% risk of failure.
Want to Learn More?
If you enjoyed this post, check out the full episode on Apple Podcasts, Spotify, or YouTube.
For more startup stories like this in emerging verticals (Web3, HealthTech, FinTech, AI/ML), follow us on Twitter @TheBreakoutsPod and follow the makers of the pod: Akkshay Khoslaa & Gokul Kumarresen.
