ecosystem / companies / metaspectral
Metaspectral logo

Metaspectral

Metaspectral is a Vancouver company that reads hyperspectral imagery in real time to identify what materials are — used to sort plastics on moving recycling lines and to analyse satellite data. It began in 2018 as PixelDrive, the machine-learning photo compressor co-founders Migel Tissera and Francis Doumet demoed at TechVancouver.

AIVancouverFounded 2018metaspectral.com
$4.7Mseed round, November 2022, from SOMA Capital, Acequia Capital, the Government of Canada and angels including Jude Gomila and Alan RutledgeTechCrunch
~5%of original file size — an 11 MB photo compressed to 583 KB live on stage, with no loss of resolutionTechVancouver stage 2018-10
1 m/sconveyor speed the sorting system reads plastics at, distinguishing grades that look identical to the eyeTechCrunch
2018founded by Migel Tissera and Francis Doumet as PixelDrive; renamed Compression.ai in 2019, then Metaspectral in 2020University of Adelaide Innovation & Collaboration Centre

The people

FD
Francis DoumetCo-founder & CEOprofile coming
MT
Migel TisseraCo-founder & CTOprofile coming

What they build

Clarity RecoverIdentifies and sorts materials on recycling lines in real timeProduct page ↗
Clarity GeoMineral mapping and crop monitoring from earth-observation imageryProduct page ↗
Clarity PlumeEmissions detection and tracking from spectral dataProduct page ↗
PixelDriveThe original 2018 product: machine-learning photo compression, later opened to developers as the Compression.ai API

Open roles via the TechTO Job Board

Working at Metaspectral

Browse open roles across the community on the TechTO Job Board — Canada’s most connected tech network. Metaspectral’s roles show up here whenever they post.

Open roles on the TechTO Job Board ↗Or their own careers page: metaspectral.com/careers ↗

Jobs live on jobs.techto.org — the only structured source. We link to it rather than scraping career sites, so the board is always the source of truth.

On the TechVancouver stage

Every talk is searchable — ask the archive about Metaspectral

About

Metaspectral started life as PixelDrive. In October 2018 Migel Tissera and Francis Doumet took it to the TechVancouver stage and dragged photos into a browser: a convolutional network trained in TensorFlow compressed an 11 MB image to 583 KB — about 5% — with no loss of resolution, and the pair explained that the higher the resolution, the higher the compression factor. The company behind it was MLVX Technologies, for "machine learning visual experience"; Tissera's compression work came out of his PhD, which tailored compression to the specific type of input rather than using a generic formula. Video was the eventual target, photos were the tractable first step, and on stage they said an enterprise API was next and that they were hiring for sales. In 2019 the company rebranded to Compression.ai and opened the platform to developers through that API. In 2020, after NVIDIA called it a real-life Pied Piper, the defence and space sectors came asking about multispectral and hyperspectral data, and the company became Metaspectral. It raised a $4.7M seed round in November 2022 from SOMA Capital, Acequia Capital, the Government of Canada and angels including Jude Gomila and Alan Rutledge, and now sells Clarity, which identifies materials and their composition from spectral imagery for recycling recovery, earth observation and emissions monitoring.

Backers

SOMA Capital, Acequia Capital, the Government of Canada, and angels including Jude Gomila and Alan Rutledge · $4.7M seed (Nov 2022)

Quick answers

What does Metaspectral do?

It turns hyperspectral imagery into material identification — reading what something is made of, and its composition and defects, from light outside what a normal camera or the human eye captures. Its Clarity products cover recycling recovery, earth observation and emissions monitoring.

What was PixelDrive?

Metaspectral's first product, demoed at TechVancouver in October 2018: drag a photo into the browser and a neural network compressed it to around 5% of its size without losing resolution. It became the Compression.ai developer API in 2019 before the company pivoted to spectral data.

Who founded it?

Migel Tissera and Francis Doumet, in 2018. Tissera, the CTO, did his PhD on neural networks for real-time edge computing; Doumet, the CEO, has an MBA from Wharton and a master's in engineering from Stanford.

How much has Metaspectral raised?

A $4.7M seed round announced in November 2022, from SOMA Capital, Acequia Capital, the Government of Canada and angel investors including Jude Gomila and Alan Rutledge.

Suggest an edit
Work here? Claim this page
Know someone who belongs here? Suggest a profile
Maintained by TechTO · facts sourced and dated · last reviewed Aug 10, 2026