A visual, modular workflow canvas where you assemble a lab process from building blocks — and the experiment design automatically wires every module.
Every lab process is different, so Halffield doesn’t ship one rigid pipeline — it ships building blocks. Drag modules onto a canvas (entity selection, materials, plate design, sample prep, formulation, instruments, QC gates, analysis, reporting, human steps, routers, and more), wire them together, and run.
The twist is the ambient design layer. You author the experiment’s variables, endpoints, controls, and DOE factors once — and every module reads from that design automatically. Declare an endpoint and the QC gate, the report, and the scorer already target it; set your factorial and the plate layout and the actual sample count follow it. No re-entering the same parameters five times. No drift.
27+ module types with edge validation, auto-layout, and per-module configuration drawers.
The assay design is the schema the whole workflow is typed against — modules are parameterized by it, not fed over an edge.
A real manifest engine (conditions × replicates × count-rules) drives both the plate paint and the run’s sample minting.
An automatic engine for in-silico / compute modules, and a wet-lab path that dispatches human tasks — from one canvas.
Module presets, workflow templates, and canonical research archetypes (HTS, in-silico design) to clone and adapt.
Add a new capability — a new instrument type, a new assay readout, a new compute tool — by adding a module and a design binding, not by rebuilding the pipeline.