01
Compute where you work.
AI classification on your own GPU, with a streaming pipeline designed to work through large projects in manageable tiles.
LOCAL INFERENCE A NEW DIRECTION FOR POINT CLOUDS
Your points.
Your processing.
Your control.
Meet FT One. Local AI point-cloud classification, designed to put the power of your workstation to work—and keep your data in your hands.
In development / Local first. Cloud optional.
✛ A DIFFERENT POINT OF VIEW
Illustrative data · not live AI
GroundVegetationBuildings
DESIGNED AROUND YOUR WORK
Local processing + Protected edits + Open point-cloud formats + Optional cloud
01 / THE APPROACH
Big datasets.
Closer to home.
FT One starts with a simple idea: the machine on your desk should be at the center of your point-cloud workflow. Here’s what we’re designing toward.
02
Your edits deserve respect.
Protected classes and confidence-aware review, so automated classification can build on the work you’ve already done.
HUMAN CONTROL03
Cloud, when you choose.
A planned opt-in path for cloud jobs and sharing. Local processing remains the foundation of the product.
OPTIONAL BY DESIGN02 / THE WORKFLOW
From raw capture
to a clearer picture.
One connected workflow is the goal. Bring in your points, classify with context, review the details, and move your work forward.
Bring your world in.
Planned LAS, LAZ, and COPC workflows, with tiled processing for dense captures.
Let local AI assist.
Start with ground and multi-class classification, with specialized models on the roadmap.
Keep the final say.
Review confidence, protect existing classes, and refine the output on your terms.
Take your work onward.
Planned LAZ and COPC output with model provenance for a traceable handoff.
Proposed product workflow. Capabilities are in development and may evolve.
03 / WHAT’S AHEAD
Built from the
ground up.
The roadmap begins with the essentials, then expands into more specialized work.
PRODUCT VISION · NO RELEASE DATE ANNOUNCEDA local foundation
Ground classification, tiled inference, class protection, and confidence-aware output.
More context. More classes.
Multi-class models and specialized forestry and utility workflows.
Connect on your terms
Optional cloud processing, Snoopy web sharing, and an opt-in assistant.
04 / GOOD QUESTIONS
A little more clarity.
Can I download FT One today?
Not yet. FT One is a planned software product in development. This site introduces its direction; release timing and early-access availability have not been announced.
Will my point clouds need to leave my computer?
The proposed architecture runs classification locally by default. Cloud processing and feedback sharing are planned as explicit opt-in choices. Final data-handling details will be published with the product.
What hardware will I need?
The design targets GPU-accelerated desktop processing, with a CPU fallback under consideration. Supported hardware and minimum requirements will be published after benchmarking and validation.
What will FT One cost?
Pricing and licensing are still being developed. No subscriptions, processing credits, or software licenses are being sold through this website.
Is the point-cloud preview a working classifier?
No. It is a synthetic illustration of a classified landscape. The color controls demonstrate the visual concept and do not run AI or process uploaded data.
THE NEXT CHAPTER IN POINT-CLOUD PROCESSING
More intelligence.
Closer to your data.
Take another look ↑FT One is taking shape. Product and launch details will appear here.