Mila — Quebec AI Institute and the DISA agrifood project

NA-Canada (Quebec / Montréal)

Content

Mila (Quebec Artificial Intelligence Institute) is a research institute in Montreal, founded in 2017 by Yoshua Bengio. Approximately 1,100 researchers, ~1,000 students. Anchored at Université de Montréal, with McGill University, HEC Montréal, and Polytechnique Montréal as partner institutions. Quebec provincial funding of $36M (announced 2026).

DISA — Data-driven Insights for Sustainable Agriculture. Mila’s principal agrifood applied project. Uses satellite imagery and algorithms to:

• Assess existing agricultural practices
• Promote regenerative agriculture
• Sequester carbon
• Reduce soil erosion
• Reduce water pollution

Other agrifood-adjacent Mila applied projects:

• Antenna — interdisciplinary platform for uploading, classifying, and analyzing in-the-wild images of invertebrates. Advances research and conservation.
• FLAIR (First Languages AI Reality) — Indigenous languages, ASR. Adjacent to agrifood through language preservation in Indigenous communities.

The structural significance. Mila is the largest academic AI cluster in Canada and one of the largest in the world. Its agrifood work spans DISA (Rwandan-deployment + Canadian-research-pipelined), Innovasea (Canadian aquaculture), and IVADO R3AI (positioning), with portfolio coverage across Mila’s applied-projects area. The provincial $36M funding is the strongest direct Quebec government commitment to AI in any sector.

What this unit is doing in the taxonomy

Anchors the Quebec academic AI cluster with a deployed agrifood project (DISA). First unit in the field guide that represents a research institute’s deployed agrifood AI project at scale. Distinguishes from:

• Olds College Smart Farm (olds-college-smart-farm.md) — academic applied research but at a college / 3,600-acre scale.
• EMILI Innovation Farms (emili-innovation-farms.md) — industry-led nonprofit, Manitoba scale, AIVA hub.
• FCC Root AI (root-ai.md) — extension LLM, FCC Crown corp, deployment not research.

Why it matters for talks

• Mila’s DISA project is the strongest Canadian-anchored agrifood AI research-and-deployment work. Worth knowing for any talk about Canadian AI capacity in agrifood.
• The connection between Mila’s academic AI strength and the federal AI for All strategy is structurally important — Mila is positioned as the Quebec-side implementation partner.
• The Indigenous languages work (FLAIR) is adjacent to the IDSov / Indigenous data sovereignty critical voice in the field guide. Worth surfacing because it shows Canadian AI institutions engaging directly with Indigenous data and language work.

Critical context

• Mila’s agrifood deployment scale (DISA) is not yet at the level of the field’s commercial deployments. The project is real and substantive, but operational acreage / partner-farm count is not surfaced in our sources.
• The DISA project’s regenerative agriculture framing is substantively different from the productivity / yield framing of most agritech AI. Worth surfacing.
• The Mila–IVADO–Quebec positioning in AI for All is recent (June 2026) and the implementation outcomes are forward-looking.