Designing the Future of Development: An AI-Powered SDLC Learning Strategy.
In a rapidly evolving landscape of Generative AI, our software engineering teams had the tools (GitHub Copilot) but lacked a roadmap for true integration. I stepped into a high-stakes, ambiguous environment where a training pilot for AI-assisted development was stalled. With no clear scope and a team of external contractors waiting direction, I had to define the “why” and “how” of AI adoption under a tight timeline.
-
I treated the pilot like a product launch. First, I facilitated collaborative sessions with engineering leadership for problem articulation, alignment on requirements and learning outcomes and to map out “Jobs to Be Done” across the development lifecycle of the SDLC.
Using Bloom’s Taxonomy, I defined a learning outcomes and objectives framework to lead the development of a comprehensive curriculum covering everything from debugging, unit testing to Docker artifact creation and systems integration testing.
Beyond the technical, I focused on the human element. I mentored a junior Project Manager and guided a Software Engineer through their first instructional design experience, ensuring the project remained lean and agile despite the lack of internal resources.
-
I designed a rigorous feedback ecosystem to track the shift in participant sentiment and technical mastery. Through a mix of pre/post assessment comparison data, 1:1 interviews, and a 90-day integration and impact survey, I analyzed the program’s effect on core engineering pillars like testing and code quality.
By utilizing AI to assist in synthesizing these complex data points, I transformed raw feedback into actionable insights and provided evidence for confident decision-making by leadership.
-
The results were transformative, turning AI skeptics into power users. We didn’t just see incremental gains, we saw a fundamental shift in how work gets done.
Speed: 83% of our participants report increases or significant increases in feature writing speed and an 82% increase in unit testing.
Sentiment: 25% lift in tool confidence and 27% increase in tool trust post training.
Strategic Pivot: Based on the results, leadership adjusted plans for external vendors to scale a rollout, instead tasking our internal team to scale the program across the entire SDLC and include the Product and UX orgs.
-
Not only were we equipping our partners to win at AI-assisted development, our own development team leveraged GenAI tools (GitHub Copilot, Glean, Gemini, Miro, Microsoft Copilot) to accelerate execution to delivery.
To maximize ROI and reach, I leveraged NotebookLM to transform curriculum transcripts, lesson plans, prompting best practices and code instructions into interactive, step-by-step coding podcasts. This allowed us to scale the learning experience beyond the initial participant pool, providing a multi-modal approach to upskilling
“This course helped me understand how I can re-contextualize
coding into problems solvable through AI.”
- Staff Software Engineer