Over the past little while at Geoson, my focus has shifted from initial onboarding to actively applying what I’m learning in more practical, real-world scenarios. Building on the foundation I described in my first post, I’ve had the opportunity to deepen my understanding of Microsoft’s Power Platform, AI tools, and how these technologies come together in a production environment.
A major part of this phase has been working through several Microsoft Learn paths related to AI and application development. I’ve been exploring how AI agents are designed and how Microsoft 365 Copilot can be extended using tools like Copilot Studio and Azure. These modules helped me better understand how generative AI fits into enterprise workflows, including concepts like orchestration, agent design, and when AI adds real value versus when traditional automation is a better fit.
Alongside this, I’ve been learning more about AI Builder and how AI models can be integrated into business applications. This included understanding common use cases such as text recognition and data extraction, as well as how these models interact with Dataverse, Power Apps, and Power Automate. Seeing how AI capabilities are packaged in a way that’s accessible to app makers and business users has been especially interesting from a systems design perspective.
One of the most valuable technical areas I’ve focused on recently is Power Apps development itself. I spent time learning imperative development techniques for canvas apps, diving deeper into Power Fx and understanding the differences between declarative and imperative logic. Concepts like global variables, contextual variables, and collections became much clearer once I started using them in real applications rather than just learning about them in isolation.
That learning came together when I started working on canvas screens for the EzExpense Pro version. Being able to contribute in a production environment pushed me to think more carefully about app structure, data flow, and maintainability. I gained hands-on experience writing Power Fx for real use cases, designing screens with actual users in mind, and understanding how small design or logic decisions can impact usability and performance.
Beyond the technical skills, working in a production area has given me a better appreciation for how professional teams build, review, and iterate on solutions. I’ve learned how development decisions are influenced by business needs, how documentation and standards help keep projects scalable, and how collaboration plays a key role in delivering reliable solutions.
As I continue my co-op term, my goal is to keep building on this momentum. I want to keep strengthening my technical skills while gaining more exposure to how AI, automation, and low-code tools are used to solve real business problems. I’m excited to keep learning from the team at Geoson and to continue contributing in meaningful ways as the term progresses.