Our Vision
We've decided to spend the next decade fixing how the world discovers drugs. Not optimising one step. Not building another tool. Redesigning the system itself.
Drug discovery is a complex system with an information problem
Drug discovery is one of the most complex systems humans have built. It takes 12 years (or ~3,000,000 hours by our estimates) to bring a drug to market, with dozens of teams — from biologists and chemists to data scientists and clinicians — working on deeply interconnected questions.
There are ~10 million researchers and ~12,000 drug discovery companies in the world. Yet 80% of diseases still don't have approved treatments. The ten largest pharma companies spend over $100 billion per year on R&D combined. The output: roughly 50 new drugs approved per year. These numbers reflect something deeper than scientific difficulty. Yes, biology and chemistry are hard. But a lot of the inefficiency is systemic, not scientific.
The system is broken across three levels — and each one amplifies the next.
Fragmented execution and results. Results are scattered across different teams, tools, and systems that don't talk to each other. Decisions are made in meetings and never recorded, and when people leave, all that historical knowledge is lost. The latest technologies are not leveraged, with different teams using different software and data formats — resulting in a lot of manual work and no time left to stay current on competitors, new publications, or shifts in the field. On top of this, interdependencies between teams cause further delays.
Insight generation bottleneck. Because execution is fragmented and mostly manual, turning results into insights is slow. Scientists spend their time on admin — writing reports, transferring data between systems — instead of generating the scientific insights that actually move programmes forward. Because everything is disjointed and information is lost at every step, communication between teams introduces more friction and further delays decisions.
Strategic blindness. Because insights don't surface, decisions rely on stale data instead of connected, live evidence. Senior leaders have no real-time visibility into what programmes have learned. Programmes are delayed, never revisited, or not killed fast enough — because nobody has the complete picture to decide with confidence. The right information exists somewhere in the organisation; it's just not accessible at the point where the decision is being made.
Every inefficiency amplifies the next. Fragmented results make insight generation harder. Slow insight generation starves decision-makers of signal. Poor decisions misallocate resources, which fragments execution further. The loop reinforces itself. And today there is more need than ever to rethink drug discovery. The field has done amazing work to make it easier and cheaper to generate output — new AI models, faster assays, cheaper sequencing — but this massive increase in results adds more complexity and puts more stress on deriving insights. If you cannot distil the noise into signal, every new result expands your options, not narrows them.
What drug discovery needs is not better parts. It needs a system redesign — a system where every action amplifies the next one, where execution scales and insights connect, where the flood of output is continuously distilled from noise into signal.
Drug discovery needs its Google Maps moment
So what does a system redesign actually look like? Every major leap in a field comes from the same move: raising the level of abstraction so that people work on what actually matters, and the system handles the rest.
Programming went from punch cards to assembly to high-level languages to frameworks. At each step, execution got absorbed by the system, and humans moved up to the level of design and intent. The result wasn't just speed — it was entirely new categories of what was possible. You don't build Google on punch cards. Not because punch cards are slow, but because you can't think at that level of complexity when you're managing bits by hand.
Drug discovery hasn't had this shift. Scientists are still working at the level of execution and formatting when they should be working at the level of insight and decision-making. The most expensive resource in drug discovery — scientific judgment and taste — is underutilised. The tools have improved, but the fundamental level of abstraction hasn't changed.
Consider what happened to navigation. With physical maps, you had the data but everything else was manual. You plotted the route yourself, estimated drive times, checked for construction by asking around. The map was static, disconnected, and didn't scale — adding more paper maps or more people didn't help you navigate faster. 90% of your effort went to manual execution. 10% to actually deciding where to go.
Google Maps flipped the ratio. The system handles the routing. You choose the destination. Google Maps scales the execution — it has all the maps — but the real breakthrough is that you can work at any level of abstraction seamlessly. Zoom out and you see the entire journey: the best route, the time, the alternatives. Zoom in and the system surfaces exactly the information you need at that scale — a restaurant suggestion, a live traffic update. The information is always live, always connected, always compressed to the level that's useful right now. You can plan and decide at whatever altitude matters, and the system fills in everything below it.
Google Maps didn't make paper maps faster. It changed the entire paradigm of travelling, with people focusing on what actually matters: where do I want to go, and what do I want to do when I get there?
Drug discovery today is paper maps. Scientists have the data, but the information is static, fragmented across systems that don't talk to each other. What the field needs is its Google Maps moment: a system where you zoom out to see cross-programme patterns and strategic priorities, zoom in to see the molecular detail and experimental evidence, and at every level the information is live, connected, and compressed to exactly the abstraction you need. Where scientists stop managing information and start using it — working at the level of pattern recognition, judgment, and decision-making.
What we are building at Kiin
We call it Connected Scientific Intelligence: a platform that treats drug discovery as the information network, a platform that can scale and streamline execution while distilling noise into key signal to take decisions, a platform where the insights reach the people who need to act on them, at the right level of abstraction, at the right moment.
The vertical solutions will keep getting better, and they should. But the system that connects them — the architecture that turns individual insights into compounding knowledge, the infrastructure that lets millions of researchers actually build on each other's work — doesn't exist yet.
That's the problem we chose to fix at Kiin Bio. And we think it's the most important one in drug discovery.