Eason Development
Notes on technology, leadership, and the human side of both
Where's the beef?
I share the results of analyzing over a million financial facts going back to 2019 trying to see whether AI adoption is influencing publicly traded company records. This essay gives a research-backed reality check to the current state and economic predictions related to AI.
I caught my first loop today, sir
AI maturity means moving humans farther and farther from the code and process of deploying changes. I built my first hands-off loop that updated and deployed code to production. Sharing some lessons learned along the way and time-saving ideas on how to use Claude Workflows.
Post change validation... for heart surgery?
Honestly, y'all have zero excuses after reading this for why you can't test changes in production safely.
My personal health assistant
An AI powered advisor that runs on my laptop and knows everything about my medical history. Sharing some thoughts about how it works, and what combining all that data has done for me.
Tell your AI coding tools what's next
I tested three different approaches for creating implementation plans with Claude Code to see how different methods for providing context affected the content and quality of the plans. The results surprised me.
Supercharging the User Feedback Cycle
Combining in-app feedback with Claude Code skills can increase user engagement and help you build what your users need.
The Product Expert
What might existence be like for an AI agent?
Economics of Tertiary Service Providers
SaaS companies that profit by providing services that were slightly too annoying or expensive to build are in real trouble. This has major implications for Private Equity.
Synthetic UX Testing with AI Agents
Simulating users with AI agents gives you a way to get user feedback about your service. What you end up learning could be much more valuable than what you expected to learn.
Modernizing Legacy Code: The Challenge of Forgetting
Organizations develop a shared memory space that morphs and changes over time, just like an individual's memories. I discuss the theory and its implications on things like turnover and modernization projects. I also demonstrate the theory using source code examples. Finally, we ponder how LLMs might or might not change the fundamental problem.