Life-changing products in days, not decades

Since the 1990s we've been promised a bio-revolution. We looked forward to cell therapies that reprogram your immune system to hunt cancer, gene therapies that make the blind see. We knew the science for these advancements did exist. Yet, thirty years later, we are still waiting.

The bottleneck, it turns out, has not been in discovery, but in the brutal, decade-long journey from breakthrough to patient. Penicillin was discovered in 1928, and took 15 years to reach mass production. A century on, timescales seem to have collapsed only marginally. Meanwhile, valuable breakthroughs are lost forever, and patients run out of time. This, we think, is a betrayal of human progress.

A batch size of one

So why does development still take so long? The old model of pharmaceutical research and the accompanying apparatus of regulatory frameworks, manufacturing systems, and quality processes was built for mass production. A drug was manufactured for millions of patients, meaning you only had to validate one process, from which you could manufacture millions of identical doses.

But in modern biology, products are highly individualised and manufacturing variability is huge. Each batch is characterised by trial-like uncertainty, scrutiny and risk. Validation doesn't scale with volume. The problem is that tooling hasn't adapted to this new reality.

Friction is not rigour

We started this company building hardware for real-time monitoring of cell cultures, thinking more data would solve the problem.

But we were wrong.

Knowing what happened in the bioreactor didn't matter if we couldn't track who operated it, what media was used, which cell line was selected, or what had happened upstream six weeks ago that might explain why a current batch was failing. The answer didn't exist in any single system. It had to be manually assembled piece by piece from paper records and disconnected databases. This made it very hard to find grounded, verifiable answers. A simple question—what batch was this? who approved it?—could take hours to answer. Every deviation would trigger a cascade of manual documentation, and every change would require months of re-validation. To make things worse, none of this effort made products safer.

The truth is that soul-crushing paperwork, endless spreadsheets, patchworked and siloed systems, and validation documents that take longer to write than the software they describe, doesn't actually mean better traceability, process control, or auditability. Friction does not represent rigour.

The GxP wall

I spent years building flexible data systems to fix this gap and saw rapid adoption in research. But when we hit the 'GxP Wall' of manufacturing, validation and patient delivery, we reached an impasse. The same flexibility that made the system powerful in research made it terrifying for quality teams. How do you validate something that changes every week? How do you audit a system where users can modify their own workflows?

Handing off from R&D to Manufacturing meant rewriting everything from scratch. Workflows changed so fast that nobody invested in digital systems, knowing they'd be obsolete in months. So they stuck to paper. And in a way this was rational behaviour, given the lack of appropriate tools.

The execution platform

What is needed is a new foundation: a unified execution platform that serves as a single system-of-record across the entire lifecycle, from discovery through to the patient. It needs to be both GxP-native and open by design.

In Seal, every change is recorded, every action is tracked automatically, and audit trails are created as work happens. This allows scientists to build and improve their own processes and to turn their knowledge into repeatable, automated workflows.

A platform built on these principles creates a new operating model where processes run in a precise, repeatable, and predictable way, exactly as they were defined. Version control and centralised reviews make process changes safe and decisions visible. All of this allows for flexibility for scientists and operators without sacrificing regulator and management trust or patient safety.

From decades to days

Using legacy software projects can drag on for years, with unpredictable outcomes. Using Seal, scientists can digitise a process and see results in days. Our software creates a perfect digital twin of each unit operation, including the process, science, equipment, materials, and biology itself. AI models trained on this twin simulate every outcome, turning months of physical trial and error into an afternoon of computation. Changes to a workflow made on a Monday can be in production on Tuesday. By eliminating the friction of scattered paperwork and siloed operations we compress decades of work into days.

Finally there is an execution system for the bio-tech revolution, and its name is Seal.

This is how we will build the future—in days, not decades.