The current software development model in Quebec's and Canada's public sector relies heavily on external firms. The majority of public organizations have limited internal capacity to design, develop, and maintain their own systems.
This reality is not new. But the consequences are worsening as digital becomes central to the delivery of public services. The good news: a movement toward insourcing IT capabilities is underway. The critical question: how do you ensure that code produced by new internal teams meets contractual requirements and quality standards?
The Challenges of Limited Internal Capacity
When internal software development capacity is insufficient, several challenges emerge beyond direct costs:
- Insufficient knowledge transfer: when projects end, institutional knowledge is not always effectively transferred. The organization progressively loses mastery of its own systems.
- Technology rigidity: without sufficient internal expertise, technology choices become difficult to reassess. Systems may evolve according to specific practices rather than open standards.
- Reduced agility: the smallest change requires a contractual process, a rate negotiation, and a mobilization delay. The organization cannot respond quickly to user needs.
- Evaluation difficulty: without internal capacity to validate the real complexity of a request, the organization lacks benchmarks to assess estimates and scope changes.
The Digital Skills Gap: A Quebec and Canadian Reality
The digital skills gap in the public sector is extensively documented in Quebec and Canada:
- The Gallant Commission report (February 2026) found that the SAAQ “lacked the internal resources required, in terms of both expertise and numbers” for a program the size of CASA, and recommends a centralized entity with the expertise needed to “reduce dependence on consultants and external firms” (Recommendation 1; our translation). The SFPQ, which welcomed the report, sees it as confirming a recurring shortage of in-house IT resources across the public service.
- As early as 2011, Quebec's Auditor General noted that external resources accounted for 52% of IT labour spending in departments and agencies, and that relying on them more and more to fill vacant strategic positions “increases the risk of losing control over project delivery” (our translation). In 2012, it added that “frequent reliance on outsourcing can, over time, lead to the stagnation of internal expertise, or even its loss” (our translation). The internal expertise deficit creates a vicious cycle of dependency.
- Federally, the Parliamentary Budget Officer estimated in January 2025 that task-based IT contractors cost 22% to 25.7% more than equivalent public servants in the four departments analyzed for 2022-23, a real economic incentive for insourcing.
Insourcing Is Underway : But the Transition Creates New Risks
The movement toward greater internal capacity is not theoretical. Public organizations are investing in recruiting, training, and equipping their teams. This is a necessary and desirable endeavor.
But this transition period creates specific risks that are often underestimated:
- New teams, immature processes: recently recruited developers do not yet have experience with the organization's specific context. Internal development processes are still being built.
- Unproven technologies: the move to new technologies, even when justified, introduces a learning curve and technical risks that only experience reveals.
- Standards still being defined: quality, security, and architecture standards are not yet stabilized. Deliverables are being produced before the reference frameworks are finalized.
- Lack of perspective: when everyone is new to the process, who has the perspective needed to identify gaps before they become problems?
Code produced by new internal teams requires independent validation just as much as code produced by external vendors. The source of the code changes; the need for verification remains the same.
Production Accelerates, Verification Must Follow
In a controlled experiment published by GitHub in 2022, 95 developers were asked to write an HTTP server in JavaScript: those using Copilot finished 55% faster. The acceleration of code production is a documented fact, but technology alone does not guarantee quality.
This means that even with powerful tools, the quality of the outcome fundamentally depends on the rigor of validation processes. The faster you produce, the more you need solid and independent verification mechanisms.
A Progressive Strategy with Validation at Every Stage
Rebuilding internal capacity is a medium- to long-term endeavor. The strategy must be progressive and realistic, and every phase must include independent technical validation:
- Phase 1 : Scoping: define measurable and verifiable requirements before building begins. Independent technical validation starts here.
- Phase 2 : Experimentation: start with low-risk projects. Validate that deliverables produced by new teams meet the defined requirements.
- Phase 3 : Capacity building: as teams take on more critical systems, independent verification becomes even more important.
- Phase 4 : Validated autonomy: the organization masters its own systems, but maintains independent validation as a permanent safety net.
V.I.A.'s Response: Validation at Every Stage of the Transition
At V.I.A. Solutions, we support this transition without substituting for the teams.
Our AI coaching helps public sector executives and development experts use AI for their own effectiveness and that of their organization, in keeping with public sector standards. The coaching is one-on-one: building the right reflexes (what to delegate to AI, to what extent, following which security practices), learning by doing on a real project and talking with an expert. It is a direct way to build the in-house expertise this transition needs, rather than renting it.
The Transparence platform frames projects entrusted to vendors: a scoping AI highlights vague areas and suggests measurable requirements, each deliverable is checked then confirmed by a double human validation before payment, and real-time dashboards reveal gaps as soon as they emerge. Every action leaves a dated, signed record that becomes the project's organizational memory.
Finally, our independent technology audit brings an outside perspective when it matters: before signing, at every milestone, or when the organization lacks the senior expertise to decide.
The future of the digital public sector will come from organizations that are more capable and more autonomous. But that autonomy is built on solid foundations, and solid foundations require independent validation, both human and technological, at every stage.
Updated on