The Advocacy Window
Canada's strict-AI public-sector procurement since January 2021 totals $172.7M across 141 contracts, 87 vendors, and 61 buyers. The market is structurally three-tiered: two outsized single-vendor anchors (Palantir $36.6M at the Ontario Provincial Police, IBM $25.5M at PSPC for RPA) account for 36% of the visible market between them; a service-firm cohort in the $5M–$16M range (MDA, Levio, Thales, IDEMIA, Crescendo, Louis Tanguay, 22nd Century, SDK Tek) is AltaML's true peer set on delivery model; and a long tail under $1M is where AltaML's 6-record footprint sits. Two structural facts shape every advocacy track that follows: (1) PSPC is the centre of gravity at $87.8M / 50.8% of national spend, but PSPC procures on behalf of TC, ECCC, NRCan, ISC, DFO, ISED, and CSA — the visible $87.8M is funneled demand from end-using departments, not PSPC consumption; (2) most expected federal AI buyers are absent from the visible award stream — DND, CRA, Statistics Canada, Health Canada, CBSA, ESDC, IRCC, Public Safety, RCMP all show $0M visible AI procurement (see Target #7 below). The advocacy work splits into competing for visible re-competes (Targets #1–#4), translating the Stage 2 EO reference into named-vendor positioning at the actual using departments (Target #5), and asking the absent buyers what's actually happening (Target #7).
AltaML's Known Footprint
| Value | Date | Buyer | Province | Title |
|---|---|---|---|---|
| $819,000 | 2025-02-06 | PSPC | AB | Stage 2 Accelerating Earth Observation Innovations (EO = satellite-imagery analytics for monitoring land, climate, infrastructure, marine, resources) |
| $100,000 | 2025-03-14 | University of Alberta | AB | Consultancy: Artificial Intelligence and Machine Learning |
| $26,000 | 2025-09-12 | British Columbia Energy Regulator | BC | RFP21026002 Accelerated AI Engagement |
| $0 (ranked SA) | 2025-08-15 | Alberta Gaming, Liquor and Cannabis Commission | AB | Automation Agency of Record (RPA bid) |
| $0 (ranked SA) | 2022-06-13 | PSPC | AB | TBIPS Standing Agreement |
| $0 (ranked SA) | 2021-12-16 | PSPC | AB | ProServices Standing Agreement |
The Stage 2 EO Innovations contract is AltaML's only signed deliverable-bearing federal AI contract at scale. Whether PSPC plans a follow-on, at what scope, and on what timeline is not visible in the awards data — that's a Signa conversation, not a fact this brief can claim.
The TBIPS and ProServices Standing Agreements (2021, 2022) are commodity procurement vehicles that virtually every federal-active services firm holds (Levio, CGI, Cofomo, Alithya, Deloitte, KPMG, Accenture, ADGA, IBM, dozens more). They allow federal departments to call up AltaML for defined task scopes without a separate RFP. They are infrastructure, not a competitive asset. Neither has surfaced a downstream call-up directly attributed to AltaML in the visible award stream.
The University of Alberta and BC Energy Regulator engagements are precedent contracts demonstrating that AltaML's Edmonton-anchored corporate footprint extends beyond PSPC into provincial and academic AI procurement.
Target Accounts
Target 1: AltaML's PSPC Stage 2 EO contract — what comes next?
Anchor: $819K PSPC contract titled "Stage 2 Accelerating Earth Observation Innovations" (Feb 2025). AltaML is the incumbent.
What the awards data tells us:
- AltaML won this Stage 2 contract at PSPC in Feb 2025 for $819K
- MDA Systems holds $16.1M / 4 federal contracts in adjacent EO + computer-vision scope (satellite-derived imaging) — the closest peer
- Other federal departments with plausible EO use cases (ECCC for climate monitoring, NRCan for forest / resource monitoring, TC for transportation, CSA for space program continuity, DFO for marine monitoring) show very thin direct strict-AI procurement footprints — see Targets #5 and #7
What the awards data does not tell us:
- Whether PSPC plans a follow-on or operational deployment for AltaML's specific Stage 2 contract
- Which using department(s) AltaML's Stage 2 work is being delivered for (PSPC procures for a department, not for itself)
- Whether adjacent EO programs at ECCC / NRCan / CSA / DFO are scoped to use the same procurement pathway
Signa angle — the questions to ask, on AltaML's behalf. The awards data shows a Stage 2 contract exists. It doesn't tell us what comes next. Signa's job is to find out.
- Which using department is actually behind AltaML's Stage 2 contract — ECCC? NRCan? CSA? Someone else? PSPC procures, but a department uses the work.
- Did the using department like the work? What's the internal verdict on what AltaML delivered?
- Is there a follow-on planned? At what scope, on what timeline, in what budget envelope?
- Are there adjacent EO programs at ECCC / NRCan / CSA / DFO where AltaML could be named in the next RFP draft, before it goes public?
- What does the using-department program lead think a Canadian-domiciled AI provider would unlock for them, that a US-domiciled platform vendor wouldn't?
Target 2: PSPC AI Source List ITQ (closing September 30 2026)
- Canada's first AI-specific federal pre-qualification list. Once on it, any federal department can call up AltaML for a defined scope of AI work without running a separate RFP — the AI-specific equivalent of an SA, but with technical-evaluation criteria specific to AI capability rather than the generic professional-services criteria of TBIPS / ProServices.
- The Stage 2 EO Innovations award is AltaML's most directly relevant credentialing input. This is the highest-leverage near-term Signa engagement on this brief: federal-vehicle qualification for AI-specific buying ahead of every Stage-3-and-beyond cycle.
Signa angle — the questions to ask, on AltaML's behalf:
- Who at PSPC is actually writing the AI Source List qualification criteria? What are they looking for?
- Is there Canadian-sourcing or domestic-AI-services preference language under consideration? If yes, can it be sharpened? If no, can it be added?
- How are Indigenous-data-sovereignty requirements being scoped, and does that advantage Canadian-domiciled providers in the evaluation matrix?
- Once AltaML is qualified, which federal departments are most likely to issue early call-ups — and how does AltaML get on their radar before the call-up specs are written?
Target 3: Ontario Provincial Police — Palantir re-compete (October 30 2026)
$36.6M / 1 contract / 21.2% market share. Procured by the Ontario Ministry of the Solicitor General for the Ontario Provincial Police (OPP). Incumbent: Palantir Technologies Canada (Foundry data-analytics platform). Expires October 30, 2026.
What makes this interesting:
- Single largest contract in the dataset (21.2% of national strict-AI spend) and the single largest pickup window in the next 18 months.
- Palantir's Canadian footprint has been visibly retreating in 2024–2025 — fewer Canadian government deployments, smaller Canadian team. The OPP contract may be one of Palantir's last major Canadian government deployments. That changes re-compete dynamics: Palantir may not aggressively defend it, opening room that wouldn't exist against a more committed incumbent.
- Foundry is sticky once installed, so full platform displacement is still a low-probability play. The realistic AltaML angle is a Canadian-sourcing carve-out — advocate for the RFP to require domestic AI/ML services delivery on top of whatever platform survives, with AltaML named as a qualified delivery partner.
Stakeholder context: Police deployments of analytics platforms draw privacy-commissioner, civil-liberties, and Indigenous-reconciliation scrutiny. Engaging that landscape upstream of the RFP shapes what the RFP says.
Signa angle — the questions to ask, on AltaML's behalf:
- How is the OPP actually using Foundry today? Which workflows? Which user groups?
- Is the OPP / Ministry of the Solicitor General happy with the deployment, or are there frustrations Signa can surface?
- Is Palantir actually planning to bid the re-compete aggressively, given their visible Canadian retreat?
- Would the Ministry consider a Canadian-sourcing carve-out — domestic AI/ML services delivery on top of the analytics platform — and if so, how does AltaML get named in the RFP draft?
- How is the Information and Privacy Commissioner of Ontario viewing OPP's analytics-platform deployment? What constraints might end up in the next RFP?
Target 4: PSPC IBM RPA re-compete (February 28 2027)
$25.5M / 1 contract / 14.8% market share. Incumbent: IBM Canada (RPA — IT Professional Services, EP920-250864A). Expires February 28, 2027.
What it actually is: RPA = Robotic Process Automation — software that automates repetitive desktop tasks (form-filling, data-entry, system-to-system data movement). The contract is a multi-year professional services tender, not a software-license deal.
Why it's contestable:
- RPA is a commodity services category. IBM's incumbency at PSPC is procurement-vehicle inertia, not technical lock-in — many other vendors deliver functionally equivalent RPA services.
- AltaML's differentiation is applied ML on top of automation — predictive ML to identify which workflows are worth automating, plus the automation layer itself. That's a margin-defensible positioning against commodity RPA-as-a-service.
Signa angle — the questions to ask, on AltaML's behalf:
- Which federal department actually uses this RPA contract? PSPC is the procurement record; the using department is the real buyer.
- What workflows does it actually automate? "RPA — IT Professional Services" is a generic title; there's a specific scope behind it Signa can find.
- How is IBM doing on the contract? Is the using department satisfied with delivery, or are there frustrations? Frustration is the wedge.
- Is there an opportunity to displace IBM with a Canadian-domiciled vendor at re-compete?
- What does the re-compete RFP draft look like, and when does it open for input from prospective bidders?
- Is the federal CIO branch (TBS-OCIO) framing this as "commodity RPA" or as "AI-augmented automation"? The framing shapes who can win.
Target 5: Federal cross-sell off the Stage 2 EO reference (the federal-department surfacing track)
This is the federal-department surfacing layer. PSPC's $87.8M is procurement-of-record, not consumption — the actual end-using departments behind those contracts are where AltaML's services fit. The departments where the Stage 2 EO reference travels most naturally:
| Department | Direct AI footprint visible | Strategic relevance | Engagement priority |
|---|---|---|---|
| Environment and Climate Change Canada (ECCC) | (procures via PSPC) | Climate-monitoring AI / EO inference is the most direct technical fit for AltaML's Stage 2 capability | High — strongest technical alignment |
| Natural Resources Canada (NRCan) | $0.30M / 2 direct + PSPC pulls | Forest, mining, resource-monitoring AI; satellite-imagery overlap with Stage 2 | High |
| Transport Canada (TC) | $4.0M / 2 direct contracts | EO + transportation-data analytics extend the Stage 2 capability directly; already a buyer | High — easiest to scale |
| Indigenous Services Canada (ISC) | (procures via PSPC) | Health and social-program predictive analytics in remote communities; Indigenous-data-sovereignty narrative | Medium — political alignment, longer cycle |
| Department of Fisheries and Oceans (DFO) | (procures via PSPC) | Marine monitoring + acoustic ML | Medium |
| Innovation, Science and Economic Development Canada (ISED) | (policy lever, not direct buyer) | Federal AI policy office (not a direct procurement buyer); engaging here shapes the language that lands at PSPC and the using departments | Critical (policy) |
| Canadian Space Agency (CSA) | (procures via PSPC + own program funds) | Direct EO program; obvious follow-on conversation if PSPC scopes one | High for EO track |
| Defence Construction Canada | $0.43M / 1 | Smaller anchor; defence-adjacent infrastructure analytics | Lower priority |
| Shared Services Canada (SSC) | $0.04M / 18 | Many small federal AI-adjacent IT services contracts; thin per-contract value | Watch only |
Signa angle — the questions to ask each department, on AltaML's behalf:
- What AI work is your department actually doing today? The awards data says $0 visible at most of these departments, but operating-budget AI work and pilot funding don't always surface as published RFPs.
- What's coming in the next 12-24 months? Where will visible AI procurement happen, and what scope?
- Have you heard of AltaML, and what would you need to see to consider naming them in an upcoming RFP?
- Where does AI procurement get funneled at your department — through PSPC, through your own IT shop, through ISED's program transfers, or through a category-management vehicle?
- Specifically for ISC and DFO: how do Indigenous- and Inuit-data-sovereignty requirements shape vendor selection, and does that advantage Canadian-domiciled providers?
Target 6: Provincial / Crown buyers (lower priority — list, not lead)
Quebec health-sector AI is the largest visible provincial AI cluster but a secondary track for AltaML, not a primary one. The dataset reveals real pickup opportunities worth listing:
- Santé Québec / Crescendo voice recognition + EHR ($7.0M, expires February 3, 2027). Largest health-sector AI contract. Predictive-intelligence-on-top-of-clinical-record positioning if pursued.
- RAMQ ($1.0M / 4 contracts). Health-card identity + clinical predictive analytics.
- Ville de Montréal — Service des TI ($3.6M / 4 contracts). Quebec municipal AI/BI dashboard build.
- Service des TI (QC) ($2.3M / 2 contracts). Quebec provincial IT shop.
Alberta provincial AI is an under-served market where AltaML's Edmonton base creates natural advantage:
- Rural Municipalities of Alberta (RMA) — 22nd Century Technologies ($5.0M). AI strategy contract; precedent worth watching.
- Alberta Electric System Operator (AESO) — SDK Tek ($5.0M). Crown utility data-management contract.
- Service Alberta and Red Tape Reduction — EY Canada ($3.3M). Provincial RPA at re-compete cycle.
- Government of Alberta Technology and Innovation — Green Light Consulting ($2.8M). Provincial AI services with re-compete in October 2026.
Other provincial / municipal AI buyers (smaller, watchlist-only): Ontario Transportation (IDEMIA biometric vision $7.2M, separate vertical), BC Energy Regulator (existing AltaML buyer), Region of Peel ($0.7M), Consumer Protection MB ($1.4M), City of Kelowna ($0.8M), Halifax Regional Municipality.
Target 7: Federal AI buyers visibly absent (the negative-signal track)
A second-order insight from the same dataset: most federal departments with plausible AI demand are entirely invisible in the visible award stream. Three hypotheses explain absence — classified / sole-source procurement, AI demand funneled through PSPC under PSPC's name, or genuinely thin AI procurement to date. Each is a legitimate Signa engagement target. "Why aren't you buying?" is itself a high-value advocacy conversation, and the answer determines whether AltaML's near-term play at that department is competitive bidding, sole-source pursuit, or category-creation upstream of an RFP.
| Department | Visible AI spend | Plausible AI use case | Signa engagement question |
|---|---|---|---|
| Department of National Defence (DND) | $0M | Predictive maintenance, drone CV, cyber AI, simulation/training ML | Largest federal IT-spend department, AI-defence is a published priority — is procurement happening through classified vehicles, or is there a real services gap? |
| Canada Revenue Agency (CRA) | $0M | Fraud detection, taxpayer risk scoring, NLP for case classification | Direct ML use case at scale — is AI being built in-house, or are services contracts coming? |
| Statistics Canada | $0M | ML for survey methodology, microdata privacy, analytics modernization | Analytics is StatCan's mandate — why is the country's analytics agency invisible in AI services procurement? |
| Health Canada / Public Health Agency | $0M | Pandemic-era predictive analytics, drug review NLP, surveillance modelling | Heavy 2020-2023 demand signals during COVID — did pandemic AI demand convert to procurement, or stay in pilot funding? |
| Canada Border Services Agency (CBSA) | $0M | Facial recognition, predictive risk scoring, eGate biometrics | Active US/EU border-AI deployment peers; CBSA modernization underway — what's the procurement strategy? |
| Employment and Social Development Canada (ESDC) / Service Canada | $0M | Social-benefits eligibility ML, fraud/error reduction, contact-centre NLP | Largest federal program-delivery department — is AI scoped, or being avoided for political/equity reasons? |
| Immigration, Refugees and Citizenship Canada (IRCC) | $0M | Visa/refugee triage ML | "Chinook" tool exists publicly; given the Chinook controversy, where is IRCC actually buying AI today, and where might it scope future buys? |
| Public Safety Canada / RCMP | $0M | Predictive policing, federal cyber ML | Federal police; the OPP at $36.6M is the visible benchmark — where is RCMP AI procurement happening? |
| CSIS / CSE | $0M | Cyber AI, signals analysis (presumed classified) | Existence is a given; visibility is not — are there unclassified services contracts AltaML can pursue? |
| Treasury Board / Privy Council Office | $0M | AI policy, government-wide model governance | The departments writing AI procurement policy don't visibly procure it — where do their own AI tools come from, and is there a domestic-sourcing preference Signa can shape upstream? |
| Veterans Affairs Canada (VAC) | $0M | Benefits processing ML, predictive disability claims | Service-delivery use case adjacent to ESDC — invisible procurement |
| Canadian Food Inspection Agency (CFIA) | $0M | Computer vision for food-safety inspection, supply-chain NLP | Direct CV use case overlapping AltaML's Stage 2 capability |
The takeaway. AltaML's $945K visible footprint is in a market where federal AI procurement is heavily concentrated at PSPC and most large departments are absent from the published award stream. Signa's advocacy work is as much about asking "why aren't you buying?" as it is about competing for visible RFPs. For each absent department, the conversation determines whether AltaML's near-term play is competitive bidding (departments procuring AI but invisibly), sole-source / pilot funding (classified or under-threshold), or category-creation upstream of an RFP (departments not yet buying AI services at scale). All three outcomes are Signa-actionable.
Incumbent Map (strict AI scope)
| Vendor | Market Share | Contracts | Dominance Pattern |
|---|---|---|---|
| Palantir Technologies Canada | 21.2% ($36.6M) | 4 | Single anchor at the Ontario Provincial Police (Foundry) + smaller federal pulls |
| IBM Canada | 14.8% ($25.5M) | 1 | Single PSPC RPA professional-services contract |
| MDA Systems | 9.3% ($16.1M) | 4 | Federal Earth-observation / computer-vision (closest peer to AltaML's Stage 2 capability) |
| Levio | 6.2% ($10.7M) | 3 | Quebec applied-ML and analytics services |
| Thales Digital Solutions | 6.0% ($10.3M) | 2 | Defence / regulated-sector AI; large per-contract |
| IDEMIA | 4.2% ($7.2M) | 1 | Ontario Transportation biometric vision |
| Systèmes Crescendo | 4.0% ($7.0M) | 1 | Santé Québec voice recognition + EHR |
| Louis Tanguay Informatique | 3.3% ($5.8M) | 1 | PSPC AI R&D (single contract, not ISC pathway) |
| 22nd Century Technologies | 2.9% ($5.0M) | 1 | Rural Municipalities of Alberta AI strategy |
| SDK Tek Services | 2.9% ($5.0M) | 1 | AESO data management |
| EY Canada | 2.0% ($3.5M) | 4 | Big-4 named-AI-only engagements |
| Experis Canada | 1.9% ($3.3M) | 1 | Federal AI chatbot + RPA development |
| AltaML | 0.55% ($945K) | 6 | Stage 2 EO at PSPC + provincial / academic precedents (the client) |
Pickup Opportunities (Next 18 Months)
| Value | Expires | Buyer | Incumbent | Signa angle |
|---|---|---|---|---|
| $36.6M | Oct 30, 2026 | Solicitor General (ON) / OPP | Palantir Foundry data analytics | Largest single re-compete; Canadian-sourcing carve-out play |
| $25.5M | Feb 28, 2027 | PSPC | IBM Canada (RPA - IT Pro Services) | Federal RPA re-compete; AI Source List qualification ahead of cycle |
| $7.0M | Feb 3, 2027 | Santé Québec | Systèmes Crescendo voice + EHR | Quebec health AI cross-sell anchor |
| $2.8M | Oct 15, 2026 | Government of Alberta — Technology and Innovation | Green Light Consulting | Provincial AI services re-compete; AltaML home-turf |
| $2.4M | Sep 19, 2027 | Ville de Montréal — Service des TI | Stantec Experts-Conseils (dashboard dev) | Quebec municipal AI build |
The two anchor pickup events in the next 18 months — Ontario Provincial Police Palantir (October 2026) and PSPC IBM RPA (February 2027) — together represent $62.1M of contestable spend. Signa's advocacy preparation should compress around these two cycles plus the AI Source List ITQ deadline of September 30 2026.
Category Dynamics (3 strategic buckets)
The market consolidates into three strategic buckets that map to where AltaML can credibly compete:
- Applied ML & Predictive Services — $63.1M / 60 contracts / 36.5% of market. Where AltaML plays. Fragmented across many service-firm vendors (MDA, Levio, Thales, Louis Tanguay, 22nd Century, SDK Tek, AltaML, plus many smaller). No platform incumbent. This is where the Canadian-sourcing advocacy narrative carries most weight because the procurement is delivery-services, not platform-licensing.
- Computer Vision & Earth Observation — $24.2M / 30 contracts / 14.1% of market. AltaML's anchor capability via the Stage 2 EO Innovations contract. IDEMIA (Ontario Transportation biometrics) and MDA Systems (federal EO) lead. AltaML has demonstrated computer-vision capability through the Stage 2 EO contract; whether that capability gets extended into a follow-on contract is one of the questions Signa would ask under Target #1.
- Data, RPA & NLP Platforms — $85.4M / 51 contracts / 49.4% of market. Platform-incumbent territory: Palantir Foundry at the OPP, IBM RPA at PSPC, UiPath at federal RPA, Crescendo voice recognition at Santé Québec. Where displacement is hardest because installed platforms have switching cost. Canadian-sourcing carve-outs (services on top of platforms) are the realistic AltaML play here, not platform displacement.
The structural observation: the platform-incumbent bucket is the largest at half the market, but the services bucket is where AltaML's revenue model fits. AltaML's Stage 2 EO contract sits in the smallest bucket, but it's their only signed federal-AI reference. Signa's advocacy work concentrates on (1) finding out whether the Stage 2 EO contract has a follow-on at PSPC and adjacent EO-active departments, (2) growing AltaML into the Applied ML bucket via the AI Source List + federal cross-sell, and (3) opportunistically inserting AltaML into platform-bucket re-competes via Canadian-sourcing carve-outs.
Methodology Notes
Numbers in this brief represent Canadian public-sector procurement awards visible in the published award record (CanadaBuys, MERX, federal portals, BCBid, SEAO, Bonfire municipal portals, NRC, provincial portals), filtered to award_year ≥ 2021, with amendment records stripped (regex match on title Amend NNN + amendment award-id -ACM_ suffix) to prevent contract-value double-counting.
Strict-AI scope is enforced via:
- Title-keyword inclusion: artificial intelligence, machine learning, predictive analytics, computer vision, NLP, RPA, named platforms, named applications.
- Vendor-name inclusion for AI-only firms (AltaML, Cohere, Palantir, DataRobot, IDEMIA, Crescendo, MDA, Mila and adjacent research-led firms).
- Exclusion of generic federal IT modernization (Digital Platform Modernization, TBIPS / ProServices generic call-ups, M365 EAs, IBM commodity SAs, CX consulting, HCM/CCRM/CX/API Lifecycle/ERP transformation, generic cloud migration, EUC/UEM, Digital Policing strategy, generic security architecture, Big-4 advisory engagements that don't name AI as a deliverable).
Data-quality fixes applied in this iteration:
- Re-attributed 9 Alberta Purchasing Connection records that came through the API with empty
buyer_namefields (RMA, AESO, Service Alberta, GoA Tech & Innovation, U of A, City of Calgary, City of Lethbridge, AGLC, Environment and Protected Areas). - Collapsed the federal Intelliport environmental-EO contract (EF928-230782) from 5 co-awardee duplicates to a single representative record at $1.15M.
- Dropped 4 UiPath records titled only "OTHER COMPUTER SRV" — too vague to attribute to a specific scope.
- Folded "Public Works and Government Services Canada" → canonical "PSPC" buyer.
Channels not visible in this dataset (and not claimed in this brief):
- Pilot funding through ISED, NRC IRAP, and similar program-transfer instruments
- Operating-budget AI services purchases by federal departments (not procured via published RFP)
- Provincial and university research-procurement (visible only when listed on national portals)
- Sole-source contracts under thresholds that don't trigger published-award reporting
- Standing-agreement call-ups against TBIPS / ProServices / similar commodity vehicles (not always reported in the public award stream — these are excluded by methodology rather than gap)
The growth vector — finding out what comes after AltaML's Stage 2 EO contract, qualifying for the AI Source List, and cross-selling AltaML's services to TC / ECCC / NRCan / ISC / DFO / CSA — is where the next 24 months of Signa's advocacy work concentrates.
Companion artefact: the AltaML AI Market Intelligence dashboard at applied-ml.dashboards.publicus.ai (password AltaML2026) carries every contract behind these numbers, drillable by vendor, organization, and category. Stage 2 EO, Palantir at the OPP, IBM PSPC RPA, Crescendo Santé Québec, MDA Systems EO, and Levio Quebec ML all surface as clickable rows with full incumbency and pickup-window context.