This summer, our newRoot Labs interns explored a big question: Which AI tool is best for you? Using Excel pairwise comparison formulas, Python web scraping, AWS storage, and Tableau dashboards, they built a system to compare LLMs across efficiency, accuracy, subject mastery, and reasoning ability. Their reflections on AI in school vs. the workplace reveal how learning with AI changes the way we build knowledge.

During my internship at Red Cedar, I discovered that the true essence of software engineering goes beyond writing code—it’s about empathy, problem-solving, and carrying complexity so others don’t have to. From organizing proposals to merging LLM datasets with Python and learning DevOps practices, I gained technical skills and human-centered insights that will shape my career as a computer engineer.

This summer, I learned how economics and technology intersect in unexpected ways. As an intern in Red Cedar’s Technology and Innovation office, I applied classroom skills to real-world challenges, from building Excel formulas to writing my first Python code. Along the way, I discovered that economics isn’t just about markets—it’s about learning how to think, solve problems, and apply that mindset to data science and innovation.

Many federal AI efforts stall at the pilot phase—looking impressive but delivering little real-world impact. This article explores why that gap exists and how agencies can shift from flashy demos to outcome-driven automation that supports audit readiness, decision-making, and mission success.