Detect everything
Touching and overlapping particles resolved, not skipped.
Every particle, measured and delivered
Size, shape and distribution from SEM and TSEM micrographs, whatever the instrument. One minute instead of thirty, and the numbers land in the systems you already run.
Manual measurement doesn't scale, and it doesn't agree with itself.
Two analysts measuring the same micrograph get two answers. Vessel applies one calibrated pipeline to every image, so a batch measured today is comparable to one measured last year, on a different instrument.
Stay in control
Preview, filter, and hand-correct before anything is committed.
Report without retyping
Statistics go to your record system, or out as CSV and XLSX.
From micrograph to reportable numbers.
01
Import a batch
Drop in a set of micrographs from any instrument, in any vendor format. Nine images or nine hundred, the run is the same.
02
Calibrate the scale
Draw along the scale bar, type what it represents, and every measurement comes out in nanometres instead of pixels. Instrument banners get cropped out of the analysis in the same step.
03
Set the model and preprocessing
Pick a segmentation model and variant, set thresholds, background subtraction and particle filters. Save it as a preset and the next batch of that sample type runs identically.
04
Review and correct
Work through the detections in the workbench: filter by size or shape, redraw a boundary, exclude an artefact. Nothing is locked until you sign it off.
05
Compare and report
Min and max Feret, AED, circularity and your own metrics, per particle and summarised per sample. Compare runs side by side, then push the numbers onward.
06
Track every batch
Queued, segmenting, awaiting review, signed off - per batch, with who submitted it and which material it was. A lab running several instruments sees the whole queue in one list.
07
Audit logging
Every change and every sign-in is appended as it happened and never edited. When an auditor asks how a number was produced, the answer is already on the record - who ran it, on which images, with which profile.
Thirty minutes by hand. One minute in Vessel.
Measuring one micrograph by hand takes about half an hour, and the result depends on who did it. Vessel measures every particle in it the same way, every time.
30×
faster per micrograph
0
variance between two runs of the same image
Bias
Judgement drifts between analysts and across a shift. Vessel applies one threshold set to every particle, so the call is made once in the profile.
Reproducibility
A hand measurement can't be repeated, only redone. Same image and profile gives the same statistics months later, and the audit log shows which profile produced them.
Three ways in. One set of results.
Measurement only pays off once the numbers reach the record. Vessel is built to be driven by something other than a person clicking - a scheduler, your LIMS, or an agent.
01
Integrates with any API
If your system has an API, Vessel writes to it - your endpoints, your schema. No middleware, no export folder to babysit.
LIMS · ELN · data warehouse · object storage
02
Our own documented API
Everything the workbench does is an endpoint. Submit images, run a profile, pull the statistics. Script it or schedule it.
REST · webhooks · batch jobs
03
MCP compatible
The same operations as MCP tools, so an agent can run a batch and act on the numbers. Scoped per tool, off until you enable it.
Agentic workflows · opt-in
Runs where your data has to stay.
The same application, two deployment models. Your security team picks; nothing about the analysis changes.
On-premise
Vessel runs inside your own network, on your own hardware. Micrographs and measurements never cross your perimeter - air-gapped installations included.
- No data leaves your infrastructure
- Works fully offline, no licence phone-home
- Your identity provider, your backup policy
Cloud
A managed deployment in your Azure, AWS or GCP tenant - or ours. Rolled out in days, scaled up when a batch is large and back down when it isn't.
- Azure, AWS or GCP - your tenant or ours
- Elastic capacity for large batches
- We handle updates and monitoring
Vendor agnostic
Mixed fleets, one pipeline, comparable results.
Same build either way
Move between on-prem and cloud without changing how you work.
Export
CSV, XLSX, or your reporting format.
Let's talk about what you need to measure.
Tell us your sample types, where Vessel has to run, and which systems it has to talk to.
Usually a reply the same day. No sales funnel.