Skip to content

ODSP2167 Pre-recorded On demand

Scale governed AI from pilots to production

Summary

From Microsoft's AI summary

CGI leaders and Microsoft's J.P. DeCuire discussed deploying GitHub Copilot in enterprises. One example was modernising 28 applications, including C++ and VB6 systems, in a 24x7 manufacturing environment, with Copilot used for code analysis, generation and testing. Another was a state government cash-flow forecasting project on Azure that extended reliable forecasting from two days to 30. Advice covered measuring adoption across the SDLC, controlling AI costs and avoiding model lock. CGI introduced its AI Adoption Accelerator Framework (A3F), with five phases over 10-16 weeks, and reported internal adoption above 90%.

Session description

Microsoft's catalogue description

Many organizations have piloted AI and deployed copilots, but scaling across the enterprise requires production-grade governance. This session shares frameworks for moving from experimentation to measurable impact: governance models, LLMOps/AgentOps, risk management, and human oversight. Attendees leave with practical guidance for turning AI investments into repeatable business outcomes.

← All sessions