I’ve always believed that good software engineering starts with simplicity.

For years, I’ve leaned toward server-side rendered applications — clean, predictable, and easy to maintain — rather than the increasingly complex, JavaScript-heavy SPA ecosystem.

Recently, as part of a personal project, I’ve been exploring how today’s AI/ML and ETL workflows can integrate naturally with modern, lightweight frontend patterns. That led me to experiment with FastAPI + HTMX + Jinja2, and I’ve been surprised by how well these tools align with my long-standing philosophy:

• FastAPI gives me a fast, clean Python backend that pairs naturally with data pipelines and machine learning models.
• HTMX lets me keep the UI server-driven, while still offering the interactivity users expect from modern apps — without the complexity of a full SPA framework.
• Jinja2 provides a simple, expressive templating layer that keeps everything cohesive.

This combination has finally pulled me away from my long-time reliance on PHP (which I still enjoy) and into a more modern, Python-based ecosystem — especially helpful as I deepen my focus on AI, machine learning, and ETL engineering.

Footnote: As always, different projects call for different approaches. This post simply reflects my own experience and preferences — I have great respect for the work happening in the JavaScript, SPA, and other tech stack communities.