Pre-launch · design-partner pilots open for 2027Synthetic defects · edge inspection · NVIDIA stackApply for the pilot

Company

Zero defects should not depend on waiting for defects

Viseena is a pre-launch company building inspection that learns from synthetic defects, so a line can catch a fault the first time it happens.

Company

Stage
Pre-launch
Pilots
2027
General availability
2028 target
Contact
hello@viseena.com
Mission

An AI eye that learns faster than the factory changes

Our vision is a world where zero defects ship: every product inspected by a system that adapts to new parts in hours, without waiting for failures to happen.

We start with surface defects on high-volume lines and plan to expand into assembly verification, metrology and robotic inspection cells.

Focus2027

Where we start

  • Surface defects on high-volume lines
  • Electronics, automotive, packaging, consumer goods
  • Design partners first, general availability after
Plans may change as pilots teach us.
Why we started

The data problem nobody could wait out

Deep-learning inspection works when it has enough labeled defects. The defects that matter most are the ones a good line almost never produces. Waiting for them means shipping escapes in the meantime.

Rendering those defects from CAD, with the labels built in, turns a waiting problem into a compute problem. That is a problem GPUs solve.

InsightRare

The costliest defects are the rarest ones.

ApproachCAD

Render them before they happen.

Principles

How we work

  1. 01

    Measure on real parts

    Synthetic data is a means. Accuracy on real parts is the result.

    Evidence
  2. 02

    Say what it cannot do

    We tell customers where a model has not been validated.

    Plain
  3. 03

    Work at the station

    We design for operators, engineers and the realities of a plant floor.

    Floor
Roadmap

What we plan to build

Draft roadmap. Dates are targets and will change with pilot feedback.

WhenMilestoneStatus
H1 2027Design-partner pilots: Synvex, Coreon, Veridex on one line eachPlanned
H2 2027Pilora defect reports and triage; Twinix station studiesPlanned
2028General availability; fleet management across plantsPlanned
2028+Assembly verification, metrology and robotic inspection cellsPlanned
Careers

Open roles

Draft list of roles we plan to hire for. Email careers@viseena.com with the role in the subject line.

Computer visionSenior engineer, sim-to-real adaptationRemote, US or EU
Synthetic dataOmniverse and rendering engineerRemote, US or EU
EdgeEmbedded ML engineer, Jetson and TensorRTHybrid, Detroit or Munich
FieldApplications engineer, inspection stationsOn site, US Midwest
ProductProduct designer, industrial softwareRemote, US or EU
Our stack

What we build with

Omniverse ReplicatorDomain-randomized, auto-labeled synthetic defect scenes
TAO ToolkitTransfer learning and per-SKU adapters
DGX / HGXSynthetic generation and model training
Jetson Orin and IGXRuggedized edge inference at the station
TensorRT and DeepStreamMulti-camera pipelines at line speed
HoloscanLow-latency camera and sensor I/O
Next step

Build inspection that learns

If you work on a line with defects you can't catch yet, or want to help build Viseena, get in touch.