The ITS Application Development team has kicked off a new AI Lunch & Learn series. Coordinated by AppDev Manager, Tom Brown, the series is designed to help staff explore AI thoughtfully balancing excitement about new tools with the practical realities of governance, security, and long-term maintainability.
Why We’re Doing This?
The series has three main goals:
- Improve prompting skills
- Boost efficiency with AI
- Create space for honesty
Recent sessions have focused on prompt engineering for developers, including:
- Context Engineering and Context Servers: How to feed AI the right information at the right time and maintain shared memory across tools, rather than relying on copy-and-paste.
- Role Assignment: Asking AI to act as a reviewer, architect, or teammate to shape tone and level of detail.
- Contextual Priming: Supplying background, goals, and constraints so responses fit our real environment.
- Reverse Prompting: Starting with the desired output and working backward to design the right input.
- Iterative Prompting: Treating AI like a conversation, refining prompts based on what works and what doesn’t.
- Vibe Coding: Comparing high-level prompts to spec-driven, agentic development, where AI follows clear requirements and structured workflows. Together, these approaches help us keep the discipline of good software engineering while using AI for exploration and acceleration.
The series has already featured several voices from inside and outside state government:
- Casey Kawamura (ITS AppDev) — Presented a product developed using vibe coding to gather and display diagnostic information from workstations used to troubleshoot a recent P2 incident.
- Phil Merrell (BSU) — Shared how Boise State University is rolling out an AI platform for students and faculty.
- Taylor Bothke (ITS CPO) — Highlighted the importance of data categorization in AI governance and risk management.
- Chris Gozzo (ITS DBA Team) — Demonstrated personal experiments running LLMs on Raspberry Pi (separate from state resources), giving a glimpse into low-cost, hands-on AI exploration.
- AWS — Presented on what is possible in the agentic arena around application discovery and modernization.
Looking ahead, upcoming speakers include:
- Eric Neely (ITS AppDev) Nov. 24th — Sharing research into MCP servers and what it means for future AI integration.
Future sessions will continue to focus on practical applications, emerging best practices, and responsible adoption. Interested in participating? Contact Tom Brown to get added to the series.
