Our technology

We’re creating a 3D reconstruction and
scene-understanding system that takes real-world spaces and makes them digital. Hyper accurate. Incredibly robust. Game changing.

How it works
01
Spatial Data Origination

We capture rich, accurate 3D datasets of real-world spaces.

Our rigorous and documented reality capture process generates Pupil’s own volumetric and image-based spatial datasets. The complete chain of custody of all of our spatial data delivers value across the property transaction lifecycle and beyond.
Pupil Spatial Data Origination
02
CLOUD

Our cloud native architecture processes the Spec and Stak datasets at scale.

Once in the cloud, proprietary computer vision algorithms turn the raw data into millimetre-accurate 3D digital reconstructions of real-world spaces. From there, we can rapidly deliver our rich datasets.
Pupil cloud native architecture
03
PUPIL AI

We’re constructing the largest verified built-world volumetric datastore there’s ever been.

It puts us in a unique place to leverage cutting-edge advances in machine learning for 3D datasets. And with Pupil AI's scene-understanding system we can automatically learn what’s most important about any space.
Pupil AI
Backed by Innovate UK

“Pupil has very innovative and potentially game-changing AI. Its first-of-a-kind nature makes it doubly attractive to UK PLC.”

— Innovate UK
Innovate UK logo
Innovate UK is part of UK Research and Innovation, working in partnership with the country’s top universities, research organisations, and companies to maximise benefit of knowledge, talent, and ideas.
Privacy and Security at Pupil
Privacy and Security

Pupil’s encryption protocols ensure that all of our spatial data is secure throughout our capture and delivery pipeline.

We deploy the highest standards of security at our application layer during every property capture, at the transit layer when data is uploaded to Pupil’s cloud, and at the server layer with Pupil’s AI and engineering systems. This means that our spatial data is secured from falling into the wrong hands.
Built on secure and scalable AWS infrastructure.
Learn more at aws.com
Interview series

Our research and data engineers are working on pioneering AI projects that underpin Pupil’s market leading products

Watch the videos below to see Pupil’s Darkroom team discuss a range of topics from machine learning models to semantic segmentation and the impact they are already having.
Divia Bhatnagar
Director of Technology
Apurv Nigam
Head of AI
Juil Sock
Research Engineer