Our Vision for AI in Pharma

At Formation Bio, we are building a tech and AI driven pharma company differentiated by radically more efficient drug development.

As AI accelerates drug discovery, generating more promising candidate drugs than ever before, we believe the most consequential bottleneck to bringing more medicines to patients is the cost and time of drug development and clinical trials. In fact, we started Formation Bio with this thesis in mind: build a pharma company that can run clinical trials and drug development more efficiently.

Potential new medicines only stand a chance to reach patients if they can advance through clinical development, and right now, the clinical development bottleneck severely limits how many drugs our industry can develop. We believe that our AI agents and domain-customized models will dramatically improve the efficiency of many aspects of drug development, and ultimately, enable more new medicines to reach patients.

Our focus on leveraging AI to transform drug development, rather than drug discovery, is less typical in our industry. This article explains why we are focused on this particular part of the biopharma value chain.

We are living in a golden age of drug discovery, however, we will be unable to capitalize on this innovation if drug development remains a bottleneck

The technology revolution across drug discovery and AI is generating more promising drug assets faster and more efficiently than ever before. However, the cost of drug development and clinical trials is only getting more expensive. As a result, the biopharma industry, which can already only advance a fraction of the drugs they discover, will find that more and more promising drugs are shelved for a lack of development bandwidth and resources.

Generative AI presents a number of new and exciting opportunities for drug discovery. Many of the AI-led discovery companies have the ability or potential to discover 25, 50, 100 drugs for the cost and time it takes a more traditional company to discover one drug candidate. However, these companies are only able to finance or develop a small fraction of the drugs they discover. Small biotechs struggle to raise capital beyond their first few clinical stage assets while large pharma have capped R&D spend (i.e., public market investors expect Pharma companies to spend no more than 20% - 25% of Sales on R&D), limiting the number of programs they can develop.

Simultaneously, while drug discovery is getting more efficient, drug development is actually getting more expensive.

There has been a considerable increase in trial costs per developed drug and a 2X increase in clinical trial timelines over the past 40 years. Today, Phase 2 clinical trials can cost $20M-$50M+ and a Phase 3 trial can cost $100M+. This cost limits the total number of drugs companies can develop.

It’s not just cost, but also time that limits our industry’s throughput. Across the standard 20-year patent life for a drug, only 1 - 1.5 years is discovery while 10 - 12 years is development. There is simply more impact to be made in drug development. Speeding up development by 1 year, not only accelerates the potential availability of that drug for patients, but also creates a valuable additional year of patent life.

As a result, many companies end up developing only a fraction of the drugs they discover and countless promising drugs end up stalled in larger pharma and smaller biotechs alike.

Our business model

At Formation Bio, we partner with biotech and pharma companies to enable more of their pipeline to be developed so that more drugs can ultimately reach patients.

We acquire, partner, or in-license clinical stage assets from pharma, research institutions, and biotechs, and provide all the capital and capabilities to get their programs through proof of concept (i.e., post Phase 2 readout) and beyond.

As a result, smaller biotechs don’t have to build out costly drug development teams and can advance more than just their lead assets. For large pharma companies, we can create an “off balance sheet" pathway for development, enabling them to take more shots on goal in a P&L efficient way.

Like others before us, we employ a Hub and Spoke model to create win-win structures with our partners. We structure each NewCo bespoke to partner preferences — which could include equity, upside, or royalties — and provide the capital and capabilities needed to advance the asset. We win when our partners win – we can sell the drug to pharma post Phase 2 readout, find a pharma partner after PoC is achieved, continue to advance the program ourselves, and even commercialize if needed.

A licensing-first approach isn’t new. Companies like Allergan pioneered an ‘external innovation’ strategy where they partnered with biotechs and discovery companies rather than building their own internal discovery efforts. More recently, Hub and Spoke models like Roivant and BridgeBio have scaled and industrialized this approach, creating value not only for their shareholders and partners, but also, and more importantly, for patients.

Our team knows these approaches well. Our Chief Development Officer, Gavin Corcoran was the former Chief Medical Officer at Allergan and we have a number of drug hunters on our team experienced in what it takes to license and acquire promising drugs in a search space where adverse selection and information asymmetry needs to be top of mind as we diligence drug programs.

Where our approach is differentiated is our ability to run drug development and clinical trials more efficiently. As a tech and AI-native company, we’ve built our own, in-house, tech driven CRO, with all the Quality and Drug Development systems, processes, and expertise needed to run clinical trials. We’ve built proprietary technology and data infrastructure that not only allows us to run trials more efficiently but also sets us up to uniquely capitalize on building and training AI models.

Since we founded the company, we’ve been able to show that we can run trials faster and more efficiently. We’ve supported hundreds of studies and have operated across the trial stack – as a recruitment vendor, technology partner, SMO/trial sites, and CRO – and our experiences across each nut and bolt of running trials give us unique insight on how to transform drug development.

While our organization is young by pharma standards, our team members are very experienced in drug development. Across our team and network of advisors, we’ve collectively been involved in 75+ regulatory submissions and 45 approved drugs.

At Formation Bio, we are continually growing our pipeline of drugs, and expanding our suite of partnerships that will help us develop and commercialize drugs most efficiently. In the coming years, we hope to partner with many more companies to provide new pathways for biotech and pharma companies to advance more new drugs to patients.

AI has the potential to transform drug development

We are excited about AI’s potential to meaningfully address the drug development bottleneck.

It doesn't take too much imagination today to see how fine-tuned and customized LLMs can streamline many pharma functions including medical writing, protocol development, biostats, and more. The regulatory function in a pharma company involves synthesizing FDA meeting minutes and publications to best anticipate what regulators may want to focus on based on precedent and ever-changing guidance. Many drug development processes today are highly manual, time intensive, and costly, and AI, coupled with the right human oversight and reinforcement, can automate many of these functions.

We have begun incorporating LLMs and AI models in everything we do. Our plans for AI innovation and investment span short, medium, and long term horizons, and always keep privacy and the best interests of patients in mind.

We are currently working on many projects with shorter term horizons that focus on workflow automation tasks – i.e., enabling 1 person to do their work 10X more efficiently. Some examples of projects include: leveraging AI to generate high-performing patient facing content and pre-screening questionnaires tailored to specific patient cohorts, using AI to assist with producing adverse event reports in minutes rather than hours, building AI-driven recruitment channel strategies that can maximize recruitment speed and patient diversity while minimizing recruitment cost, and using AI models to support optimal study design with populations and endpoints selected to maximize useful evidence generation and reduce overall costs. Fine-tuned models with human experts in the loop can significantly improve these workflows and give our team more leverage to focus on the most strategic aspects of their work.

Our verticality gives us a unique competitive advantage. In the long term, it allows us to think holistically about how pharma can be transformed with AI beyond task-centric point solutions. In this spirit, we are also working on an AI R&D Scientist that can provide decision support and eventually steer higher quality R&D decisions. Finally, to ensure we’re successful in the long term, we are also curating data sets that will allow us to build AI models to better predict toxicity, tolerability, and one day, efficacy.

We recognize that these transformative efforts will not be easy and require expertise that we may not have today. We are looking to partner with the best minds across AI, pharma, and science to realize our mission. Come join us if our mission interests you!

People, culture, and values

Since the company’s inception, it is clear that our people, culture, and values are our biggest competitive advantage. As we look to build a pharma company that will fundamentally transform how we bring new medicines to patients, we will win not because of one specific innovation or technological breakthrough. We will win because we have company culture and values that enable us to consistently innovate and execute with rigor and velocity.

AI is moving at a rapid pace and organizations that are able to move with greater urgency and velocity and balance long term missions with a focus on short term execution will ultimately win. While we have a head start today, we believe slope matters more than y-intercept. It's important for us to continue to stay curious, learn new AI tools faster, and adopt them more aggressively than our peers.

It’s likely that many AI and biotech discovery companies will be funded across the next few years, each with its unique technological and scientific approach. I believe our differentiated strategy focused on drug development will allow us to capture the benefit from this cambrian explosion in drug discovery.

The journey ahead will not be easy. Pharma is a tough space to crack, and progress will not always be linear. But it is a journey worth taking. There are few missions more important than helping humans live longer, healthy, and happier lives, and very few companies have the will, curiosity, and intensity to enact real change. Observing the ingenuity, tenacity, and dedication of our people over the past few years gives me so much confidence that our team at Formation Bio is absolutely up to the challenge.