Three weeks. That is roughly how long a batch of engineered immune cells can sit in release testing before a hospital is cleared to use it. SpectaCell, a company that came out of a semiconductor research lab, says it can answer the central question those tests ask — did these cells learn to kill? — in about four hours.
It launched on 8 October 2026 with €6 million ($6.7 million) in seed financing led by imec.xpand, with PMV's Flanders Future Tech Fund, InnovationQuarter Capital, Qbic and VIB.
The bet is narrow and testable: in cell therapy the scarce input is measurement, not biology, and a label-free, chip-based readout can become the release standard.
Cell therapies such as CAR-T and T-cell engagers are living medicines. Their potency depends on behaviour, and behaviour is difficult to measure. A batch can look correct on every surface marker and still fail to kill its target.
The conventional answer is a functional assay: mix the cells with targets, wait, and read an average. That average hides the interesting part. Two cells from the same batch can behave differently, and a slow, label-heavy assay turns a release decision into a multi-day queue.
What SpectaCell actually measures
The platform pairs effector cells with target cells inside a micro-compartmentalized cartridge, then watches what happens. Chip-based, lens-free imaging captures many interactions at once at single-cell resolution, and AI analytics converts the resulting images into kinetic and morphological readouts.
The company says a single run covers more than 10,000 cell-cell interactions. There are no fluorescent labels and no sample preparation. That matters more than it sounds. Fluorescent labels are how most assays make cells visible in the first place, and a label can change the behaviour being measured.
Cell-cell interactions measured per run
Real-time, label-free tracking at single-cell resolution instead of an average across a sample · SpectaCell, 2026
The output is a behaviour-based potency signal rather than an endpoint number. Instead of asking whether the sample killed anything, the cartridge records which cells engaged, how long contact lasted, and whether the target died. That is a different kind of dataset, and it is the raw material for the company's real product claim.
Why imec is the interesting part
imec is a Leuven-based nanoelectronics institute — the kind of place that usually shows up in stories about advanced chip packaging, not cancer immunotherapy. SpectaCell is the first spin-off from a long-running collaboration between imec, VIB and Ghent University, and its imaging stack builds on more than a decade of research across the three.
That lineage explains the design choices. Lens-free imaging, micro-fabricated cartridges and on-chip signal processing are standard toolkit items in semiconductor and photonics labs. Dropped into cell biology, they produce a measurement device that looks less like a microscope and more like an instrument.
The transfer is the whole point. Biology has spent a decade producing cell therapies faster than it has produced ways to check them. The bottleneck moved from engineering the cell to knowing whether the engineered cell works — and bottlenecks like that tend to get solved by tools from outside the field. The transistor did not come from physics departments that studied biology.
Tools, not drugs
SpectaCell has no pipeline of its own. It sells the platform, and its customers are the companies that do have pipelines: pharmaceutical and biotechnology firms, and specialist manufacturers. The seed round funds development and commercialisation, and the platform is meant to be validated on partners' own therapies.
That positioning is deliberate. A tools company avoids the binary risk of a single asset and sells into every programme that needs the measurement. It also inherits the sector's structural problem: until a named partner stakes a regulated workflow on the readout, the platform is a promising instrument with a reference list still to be written.
Breakthrough medicines need breakthrough tools. Researchers need to know much earlier which candidates deserve further investment and why.— Tony Montoye, chief executive and co-founder, SpectaCell
The founding team is not academic-only. Montoye and co-founders Veronique De Vroey, Mathieu Rabaey and Jan Van de Velde bring prior roles across life-science commercialisation, cell and gene therapy operations, bioinformatics and AI, at organisations including Johnson & Johnson and Galapagos. The CEO is also the company's public contact for partnerships, which tells you where the next milestone sits.
The four-hour potency test and the batch-release problem
A potency assay is the gate between a manufactured batch and a patient. For autologous CAR-T, where the starting material is a single patient's cells, the clock is the constraint: the batch has to be tested and released before it reaches the patient, and every day spent in testing is a day the therapy is not available. Shaving days off that wait is operational, not cosmetic.
| Question | Conventional functional assay | SpectaCell platform |
|---|---|---|
| Result time | Days | About 4 hours |
| Labels | Fluorescent labels, sample prep | Label-free, no prep |
| Resolution | Average across the sample | Single-cell, over time |
| Throughput | Limited by manual steps | 10,000+ interactions per run |
| Where it runs | Research and some QC | Lead ID, process development, manufacturing QC |
The same measurement is meant to run at three points in the development funnel. In lead identification, it lets researchers compare candidates by what they do to a target cell rather than by an aggregate signal. In process development, it shows whether a manufacturing change helped. In clinical manufacturing, it supports batch-release quality control, where a four-hour test would allow same-day release.
The demand is not hypothetical. As we wrote in October, Generate has dosed the first patient in a trial for GB-5267, an AI-designed armoured CAR-T therapy. Every new entrant like that inherits the same release-testing economics, and each one is a potential customer for a faster readout.
What is missing is proof inside a regulated process. So far no pharmaceutical or biotechnology customer has been named, and a lens-free cartridge that works in a demo has to work the same way on a manufacturing floor, with an audit trail. SpectaCell says its hybrid data engine is built for exactly that, but the claim is unproven until a partner signs off on it.
What has to be true for this to become a standard
The company is explicit that its ambition is a single platform from discovery to manufacturing. The gap between that ambition and the current state is not the imaging — it is the validation. Every release assay carries regulatory weight, and replacing one means convincing both a manufacturer and a regulator that the new number means the same thing as the old one.
Does a four-hour potency readout become the release standard?
Probability: 60% — the technical case is strong, but the tool is only half the story. The missing half is a customer willing to let a four-hour, label-free readout stand next to a multi-day assay.
Where this could land
🟢 A release standard (30%)
Consequences: SpectaCell becomes infrastructure, not a vendor, and the imec lineage turns into a durable moat.
🟡 A niche first (50%)
Consequences: a healthy Series A on partner traction, but a slower climb than the pitch implies.
🔴 Validation stalls (20%)
Consequences: the platform stays a research tool, and the four-hour claim never reaches a batch record.
Six million euros is a small round for a company with a decade of research behind it. That is the point. SpectaCell is not selling a therapy with a binary outcome; it is selling a measurement, and measurements are judged by whether someone trusts them enough to make a decision. The first named partner will be the real launch.