Build Your Own Tools: Lessons from Data-Driven Pharma

Alternatively titled: Why I Stopped Making Slides and Started Building DeckBot

Hello fellow datanistas!

Have you ever wondered if building your own tools could be the key to unlocking more creativity, efficiency, and joy in your work? This is a question I recently explored in depth at the Data-Driven Pharma event, and I wanted to share my reflections and experiences with you.

In my latest blog post, I recount how my aversion to making slides led me to create DeckBot—a tool that automates slide creation using LLMs. But more importantly, I dive into the broader theme of why data scientists, computational biologists, and software developers should learn to build their own tools. From my early days struggling with graph visualizations to building open-source packages like nxviz and pyjanitor, and later shaping internal tooling at Moderna, I share stories and lessons learned about how building tools can empower individuals and teams, foster resilience, and scale impact.

I also discuss the importance of organizational buy-in, the value of open-source culture, and how the barrier to building custom tools is lower than ever thanks to AI-assisted development. The post is peppered with insights from industry peers and practical advice for anyone looking to start their own tool-building journey.

If you're curious about DeckBot, want to see the actual slides I presented, or are looking for inspiration to start building, you'll find all the details in the full writeup.

Building your own tools isn't just about coding—it's about learning, scaling your impact, and creating a culture of empowerment. Whether you're a seasoned developer or just starting out, I hope my experiences encourage you to take that first step towards building something that makes your work (and maybe even the world) a little better.

I invite you to read the full story on my blog: Build Your Own Tools. If you find it helpful or inspiring, please consider forwarding it to friends or colleagues who might benefit from it too!

Cheers,
Eric