Assay Migration & Consulting

From bench
to chip.

Moving an established in-vitro assay onto a microfluidic format is not a porting exercise — it is a redesign. We have done it from wafer to readout, and we will do it with you.

15+ years microfluidics R&D A*STAR licensed IP Wafer to readout
The honest version

Most chip migrations fail on the parts nobody budgets for.

Not the biology. The hydrogel that will not load reproducibly. The geometry that traps a bubble at hour fourteen. The segmentation model that was trained on someone else's cell line. We have hit all of these, and we would rather you did not.

Services

Three places we add value.

Chip-based in-vitro assay design

Chip geometry, cellular interaction assays, and tumor–immune models — designed around the readout you need.

Jump to section ↓

Automation of existing in-vitro assays

Migrate an assay that already works onto a chip-based, automated system — without losing the endpoint you trust.

Jump to section ↓

Computer Vision Solutions

Segmentation, classification, and tracking pipelines trained on your imagery and validated against your manual scoring.

Jump to section ↓
01 · Assay design

Design the assay around the biology.

Chip geometry is not packaging — it is an experimental variable. A micropillar barrier is what turns immune cell recruitment from a qualitative impression into a number. Compartment spacing sets the effector-to-target ratio you can actually control. Channel geometry decides whether a gradient forms or collapses.

Chip design

Geometry, specified for manufacture

Compartments, micropillar barriers, fluidic routing, and hydrogel loading strategy — designed for injection molding from the first sketch, not retrofitted for it later.

Cellular interaction

Co-culture, made quantifiable

Systems where two or more cell types must meet, migrate, or signal across a defined barrier — with the interaction spatially resolvable rather than merely observed.

Immuno-oncology

Tumor–immune interaction models

3D spheroid systems measuring recruitment, infiltration, and cytolytic kill for drug discovery — the assay class behind our own OncoMiMIC platform.

02 · Assay migration

Your assay works. Now make it scale.

We preserve the endpoint you already trust, and gain the microenvironment, the throughput, and the automation you do not yet have. The readout is validated head-to-head against your gold standard before anything is declared a success.

01

Audit the incumbent

We run your existing assay, benchmark its variance, and record where the time and the errors actually come from.

02

Re-express on chip

The same endpoint, redesigned for a 3D compartmentalized format.

03

Automate the handling

Cell seeding, hydrogel loading, dosing, and media exchange — the steps where human hands cost you reproducibility.

04

Automate the read-out

A validated vision pipeline replaces manual scoring — and scores every run identically, forever.

03 · Computer vision

Your experiment ends in a day. Scoring takes three weeks.

A 24-hour live-cell acquisition produces thousands of frames. Scored by hand, in ImageJ, by whoever is available — the throughput of the whole experiment collapses to the throughput of one person's attention. Worse, it is scored slightly differently every time.

01Semantic segmentation — cell bodies, spheroid boundaries, compartments
02Cell classification — effector vs target, live vs apoptotic, infiltrating vs peripheral
03Tracking & spatial dynamics — migration, dwell time, contact events
04Experiment gating — pipelines that decide, not just score
Every pipeline is benchmarked head-to-head against manual scoring on your own data, with the disagreements examined rather than averaged away. A vision model that only agrees with itself is worthless.
Engagement model

Small, cheap, reversible — before big and expensive.

2 WEEKS

Scoping sprint

A fixed-fee assessment. We tell you whether chip format helps your assay, or whether it does not.

4–8 WEEKS

Feasibility

A prototype chip and a first dataset. The goal is to fail fast if it is going to fail.

3–6 MONTHS

Pilot

A validated assay on production chips, benchmarked head-to-head against your incumbent method.

ONGOING

Transfer

Chips supplied, protocols documented, your scientists trained. You own the assay.

Why us

We are not consultants who read about this.

We manufacture

A real production line

Our CEO built and scaled a semiconductor fabrication business to 1,000 wafers a month in an ISO 9001 facility. We know what makes a chip manufacturable, because we make them.

We invented

Granted, licensed IP

Our CSO is the named inventor on the core US microfluidics patent, licensed from A*STAR, with 15+ years of microfluidics R&D behind it.

We use it ourselves

OncoMiMIC is the proof

Everything we would build for you, we built for our own platform first. Our services are how we prove the stack works — and how the industry migrates toward it.

Start with a two-week scoping sprint.

Fixed fee, no commitment beyond it. We will tell you honestly whether a chip format helps your assay — including when the answer is no.

Book a scoping call