The Alberta Machine Intelligence Institute, known as Amii, launched its Health Innovation Lab on May 20 as a $10 million initiative “designed to accelerate the development and adoption of AI-driven healthcare.” Development is the easy half of that sentence. For adoption to happen, a model paid for by this program has to be installed inside a clinic that is already running at capacity, and it has to keep working after the grant term ends and the researchers move on to the next thing.
That is a construction problem more than a research problem, and the call for proposals closing Sep 28 is the first document that shows how the institute and the province plan to build it. Read in order, the page is more candid than the launch was. The province has decided to buy back clinician time first.
Key facts
- The Health Innovation Lab is a partnership between the Alberta Machine Intelligence Institute and three Government of Alberta ministries, Technology and Innovation, Primary and Preventative Health Services, and Hospital and Surgical Health Services, backed by $10 million over three years according to the Amii program page.
- Amii’s program page offers up to $50,000 over a one-year term for Discovery Sprints, up to $100,000 for Exploratory Projects and up to $300,000 for Large Projects.
- Stage 1 proposals for the Health Innovation Lab run from Aug 17 to Sep 28, 2026, with review and selection in November 2026 and notice of decision in January 2027, per Amii’s published timeline.
- Amii’s May 20, 2026 launch announcement stated that the Jenkins AI Scribe clinical documentation tool was scaling to support over 850 healthcare providers across Alberta.
What the three funding tiers actually buy
Start with the money, because the tiers tell you what kind of building this is meant to be. A Discovery Sprint is capped at $50,000 over one year and is meant for early viability work. An Exploratory Project tops out at $100,000. A Large Project can reach $300,000, and the program page reserves it for work that already shows traction, has identified or secured a healthcare system partner, and sits close to a market-ready offering.
By our arithmetic, $10 million over three years works out to roughly $3.3 million a year. If every dollar went to Large Projects, the lab would fund about 33 of them across its life. If every dollar went to sprints, it would fund 200. The real mix will land somewhere between those two numbers, and where it lands is the most useful thing any Albertan can learn about this program, because it shows if the province is paying for tests or paying for products.
The optimistic case deserves to go first, and on the evidence it is mostly right. Health AI in this country has never been short of promising lab results. What it has been short of is proof gathered inside a working health system, on real records, with a clinician’s clock running. Successful applicants here get machine learning help from Amii and, in the program’s own words, “access to administrative health data.” That access is the scarce input, far scarcer than cash, and a one-year sprint that uses it to answer one narrow question is a better purchase than a large pilot whose results never get published.
The weak point sits in the Large Project clause. A healthcare system partner is the loading dock where a finished model gets delivered into a building that never closes, and the rest of the program depends on how many of those docks exist. Integration and privacy review both happen after the grant letter arrives. Neither appears in a tier table.
Why administrative burden now sits first on the list
The second step is the order of the targets. The September call names three healthcare challenges and lists them this way, starting with administrative burden in primary care, then chronic disease prevention and management, then disparities in women’s health.
The May launch described the lab’s aims differently. It led with shorter wait times through faster diagnostics, put more accurate medical imaging second, and placed reduced administrative burden third, framed as a cost saving. Between the launch and the call, paperwork moved from third to first, and imaging dropped off the list entirely.
I think that was the right call, and the reason is measurement. A documentation tool can be checked against a stopwatch, minutes per chart note before and minutes after, on the same clinicians in the same month. An imaging model has to clear clinical validation and a regulator before it touches a patient, which is the same wall facing Alberta’s portable MRI prototype right now. Paperwork is the part of the health system where a small grant can produce a number anyone can audit.
The launch announcement also named the only tool in this family that is already in production. The Jenkins AI Scribe, a clinical documentation tool led by researchers including Amii Fellow Ross Mitchell, was described in May as scaling to over 850 healthcare providers across Alberta. That is a working supply line, and it runs straight through the category the call now puts first. The program is ordering more of what has already been delivered once.
There is a caution here that the optimists tend to skip. A scribe that works in one clinic still has to be installed in the next one, on different software, with different staff, and that second installation is where good health technology usually stalls. A tier table cannot fix that. Only the partners named in the Large Project awards can.
What the January 2027 decision list will tell Albertans
The third step is how to judge the result, and the program has made that easy by publishing its own calendar. Review runs through November, and the notice of decision is due in January 2027. When the list arrives, count the awards by tier and by challenge before reading a single project description.
My forecast is that Discovery Sprints will make up at least half of the first-round awards when decisions land in January 2027, and that administrative burden projects will outnumber the other two challenges combined. If both hold, the lab is buying evidence, and if Large Projects take most of the dollars instead, it has become a procurement channel with a research label on the door.
