How Follett Software's Investment in AI Training Became a Morale & Productivity Win

Follett Software provides technology solutions that serve two-thirds of all schools in North America, with over 4 million daily logins. Behind their platform is a large engineering organization spanning multiple teams, working on mission-critical systems that schools rely on every day.

Follett Software had already deployed AI coding tools across engineering. Copilot was available. Developers were experimenting. But adoption was uneven, and the energy behind it was scattered. Some teams were pushing ahead; others were dabbling without clear direction. The organization wanted to accelerate AI adoption, but more importantly, they wanted to invest in their development team.

Up to that point, AI tooling had been a "figure it out, spend some time, do some online trainings" situation. Follett Software's leadership recognized that wasn't enough. They needed structured, hands-on enablement that would build real capability, and they needed it to land with their engineers as something more than another initiative. They needed it to feel like a genuine commitment.

DevClarity partnered with Follett Software on an AI Coding Jumpstart, a structured engagement combining foundational training for the full engineering team with hands-on pair programming sessions that put the training into practice on real work. The goal: build skills that stick, and build momentum that spreads.

DevClarity began with foundational AI coding training for the entire engineering team, covering best practices, prompt engineering techniques, and testing workflows. Then, DevClarity engineers worked alongside the team through intensive pair programming sessions focused on their active modernization work.

The engagement centered on four key elements:

• • Full-team foundational training. Equipped the full engineering organization with AI coding best practices and systematic approaches, giving everyone the same foundation regardless of prior experience.
• • Hands-on pair programming. Worked directly inside their codebase on real migration tasks, translating classroom concepts into practiced habits that developers could repeat independently.
• • Reusable workflow creation. Built a repeatable sequence for converting legacy code and creating living context documentation that the team now owns and continues to refine, proving the approach works at scale in their complex multi-project architecture.
• • Build validation integration. Configured their AI coding tool to run inside the development environment with build access, creating a self-correcting feedback loop that catches errors early and fixes them. This builds developer trust in AI-generated code.

The engagement delivered strong productivity outcomes: +48% velocity on repetitive tasks, +27% overall weekly output, and AI coding time that jumped 1.5x. The share of developers spending over half their coding time with AI grew from 19% to 47%.

Across the organization, leaders and developers alike described something that metrics don't fully capture. There was genuine excitement and organic momentum.

The enthusiasm showed up everywhere:

• • Organic champions emerged. One developer independently built a custom AI workflow to ingest server logs and produce HTML output identifying problematic nodes, extending AI usage beyond code generation into operational monitoring. Leadership described him as "just taking it to the next level."
• • Cross-team interest multiplied. A modernization knowledge-sharing session drew participants from across the organization, with one attendee from another team noting, "This is very cool... seeing it pick the legacy code apart like that was pretty impressive."
• • Developers experienced a mindset shift. A developer who participated in the pair programming sessions, called it "a great experience" while another described it as a fundamental change in how he approaches his tools.

While the engagement began with training, it turned into something that left the team energized and wanting more.

What Follett Software received. Foundational AI coding training for the full engineering organization, hands-on pair programming that produced a reusable migration workflow, a living knowledge base committed to repos, and build validation integration. The result: measurable productivity gains and daily AI adoption across nearly half the development team. More than that, it's left a team that's sharing ideas organically and already extending AI into novel territory on their own.