ChemT Biotechnology is building a “virtual cell” platform that lets scientists model and control how living cells behave, so biologics can be made faster, cheaper, and more reliably at scale. Based in Singapore, the company combines AI in biomanufacturing to give teams tunable control over cells such as T cells and CHO cells, with applications ranging from cell therapies and biologics to next‑generation vaccines and regenerative medicine.

In this spotlight, Jie Sun (CEO and Co‑Founder) and Ling Wu (President and Co-Founder of ChemT), walks through the company’s journey, what a “virtual cell” really is, and how programmable cell behavior could reshape the economics of modern therapies.​

Rethinking How We Work with Living Cells

Our conversation with Jie began not with a slide deck, but with an analogy.

“I often say cells are like babies – they’re sensitive, and incredibly hard to control. You can’t just force them to behave the way you want; you have to guide them carefully.”

In today’s biotech pipelines, companies ask cells to do extraordinary things – produce complex biologics, expand therapeutic T cells, generate viral vectors – yet still rely heavily on slow, trial‑and‑error optimisation.​

That mismatch is what pushed her and Ling to start ChemT. They wanted precise, tunable control over cell behaviour, so scientists could intentionally guide cells instead of nudging them blindly. If they can predict how cells will respond before running months of experiments with high expenses and shrinking timelines, more therapies can reach patients when they are actually needed, not years later.

“The real impact is enabling more therapies to reach patients – faster, more reliably, and at a cost that makes them accessible, not just aspirational.”​

Turning Cell Culture into a Designed Process

When asked what she wants customers to experience when using ChemT’s product, Jie said it’s confidence and leverage. Confidence that they would not be stuck in slow optimisation loops, and leverage because ChemT’s model is designed to be seamlessly integrated into their R&D work.​

In practice, that usually starts with AI‑designed small molecules.

“These are the things you can hold in your hand, put into a flask, and see the difference in the results”.

Those molecules slot into existing process workflows and deliver immediate, measurable improvements—better yields, higher viability, or more consistent cell quality. Once teams see those numbers shift, they become much more open to engaging with ChemT’s broader platform through licensing or SaaS access to the virtual cell.​

Inside ChemT’s “Virtual Cell”

“At the heart of our company is what we call a ‘virtual cell” she explained. “It doesn’t just analyse data; it models how cells respond to specific molecular changes.” Instead of opaque predictions, ChemT focuses on rational insights that scientists can actually act on, which is what allows the team to design small molecules that consistently improve performance in real manufacturing settings.​

The platform is built to work across many cell types and applications. Today, ChemT supports:​

• T cells for cell therapy manufacturing
• CHO cells for monoclonal antibodies and therapeutic proteins

and is extending into:

• NK and B cells for next‑generation immunotherapies
• HEK293 and other producer cells for viral vectors and vaccines
• iPSCs and stem cells for regenerative medicine

“The underlying principle is simple,” she said. “If a process relies on cells to manufacture a biological product, our virtual cell platform can be applied.” Looking ahead, Jie and Ling see the same logic powering use cases in agriculture, synthetic biology, and even bio‑energy – anywhere cells are tiny factories whose behaviour needs to be steered, not guessed at.​

Lessons from Turning Cell Science into Products

Jie’s advice to other founders sounds deceptively straightforward. “Validate the problem obsessively with real customers before scaling the technology,” she said. Even a big idea like digitising and programming cells has to start with one painful, concrete problem in a lab or plant. ChemT spent its early months less in conference rooms and more in manufacturing suites, listening to operators talk about inconsistent runs, fragile cell lines, and batches that worked “for reasons nobody could quite explain.”​

“Those conversations kept us honest,” Ling said.

“It’s very tempting to stay at the whiteboard level, especially when you’re excited about the science. But the real test is whether a process engineer will call you back after the first experiment and say, ‘We want to try this on another line.’”​

That discipline has also shaped the company’s focus. While the virtual cell could theoretically be applied everywhere, ChemT has chosen to prioritise high‑impact verticals where manufacturing costs and variability are huge constraints. “Ambition is important,” Ling said, “but focus is what turns ambition into results.”​

The Future of Programmable Cells

Talk to Jie or Ling for long enough and the phrase “GMP‑ready” comes up more than once. Right now, the main bottleneck is moving from promising pilots to fully standardised, regulated adoption. Many customers are enthusiastic about trying ChemT’s molecules and models, but turning them into integral parts of GMP manufacturing requires a higher bar for documentation, quality systems, and regulatory alignment.​

“We’re investing a lot behind the scenes in GMP production and regulatory preparedness,” she explained. “It’s not glamorous work, but it’s what turns a cool technology into something you can put into an FDA filing.” At the same time, the team is tightening its licensing and SaaS offerings so enterprise customers have clear paths from small pilots to long‑term platform relationships.​

Looking to the next year, Jie summarised ChemT’s priorities in three bullets:​

• Expand adoption beyond pilots, with more licensing deals and deeper platform integrations.
• Prove the platform across additional cell types and modalities.
• Explore applications outside classical biologics, such as synthetic biology and agriculture, where controllable cell behaviour can unlock entirely new products and business models.

She smiled when asked how she’ll know ChemT is really working. “For me, it’s the day when AI is building the industrial operating layer for the process engineering team. That’s when we know the idea has truly landed.”​