Building a business around MultiTherm

Our earliest ambition went beyond temperature hardware: we wanted to link every experiment’s inputs and conditions to its outcomes so accumulated data could help recommend the next experiment. I mapped out that proposed customer workflow while a machine-learning specialist explored concepts for recommending assay conditions.

Working with our advisers, I made the call to build the multi-program thermal cycler first. Running independent protocols at the same time solved an immediate bench problem within our team’s budget, whereas laboratories did not yet have the structured starting data a learning service required.

With our software engineers, we built MultiTherm’s browser portal so researchers could create temperature profiles, set step temperatures and hold times, assign different programs visually across the 96-well grid, reuse shared team protocols, and send runs straight to the instrument—keeping the proposed DataTherm recommendation service as a later stage on the roadmap.

Putting hand-built prototypes into external pilot labs and engineering a repeatable commercial instrument required very different supplier and capital commitments. Working with our engineers and finance colleagues, I built an operating plan sequencing prototype rework, calibration fixtures, and control software ahead of heavier production engineering and hard tooling.

Our operating plan tied outside production-engineering and tooling commitments to evaluation milestones and staged funding, while allowing internal engineers to explore early concepts for a second instrument without taking on external spend ahead of MultiTherm’s launch. While later commercial manufacturing still required additional capital, that discipline let our seed-stage team put working instruments and web software directly into prospective customers’ laboratories.