In a packed auditorium at PyCon India 2018 in Hyderabad International Convention Centre, Travis Oliphant stood before hundreds of developers and made a confession that excited me a lot. Here was the creator of NumPy—a library that powers everything from physics simulations to machine learning models used by millions—and he said this: “I’m actually more proud of my daughter than all the other work I’ve done.”
He had just spent twenty minutes walking us through his impact on scientific computing. SciPy. NumPy. Anaconda. Tools that literally changed the course of human events, as he put it. And yet, when his daughter wrote him a letter saying “Dad, I really appreciate you,” that moment meant more to him than any of it.
This wasn’t false modesty. It was a fundamental principle that had sustained him through two decades of open source contribution: people first, technology second.
The Beginning: A Graduate Student’s Passion Project
Travis Oliphant didn’t set out to change the world. He was just a young graduate student at the Mayo Clinic trying to solve a frustrating problem. He was working on MRI and ultrasound imaging, and he needed to get medical imaging data into Python memory—fast. But everything available at the time was painfully slow, converting data into inefficient lists of floats that couldn’t handle gigabytes of information.
There was already an array object in Python called Numeric, created by Jim Huguenin. Travis didn’t know how to write Python extensions yet, but he found code from Michael Miller and an essay by Guido van Rossum about reference counting. He studied them. He learned. And then he started building.
What came next was SciPy—a collection of scientific computing libraries he released in 2001. But here’s where the story gets interesting: Travis’s wife didn’t realize he was delaying his graduation to work on this. They were living on very little, she was expecting him to finish, get a job, and actually pay for things. But Travis kept writing code late at night, driven by something more powerful than a paycheck.
It was the feedback loop. Someone across the world would use his code and write back: “Thanks, this is great. Maybe you could add this feature?” That connection of helping someone he’d never met solve a real problem—became addictive in the best possible way.
From One Graduate Student to Millions: How NumPy Changed the World
By the mid-2000s, the Python scientific community faced a crisis. There was a split happening between Numeric and a newer array object called Numarray. Travis watched, concerned that his beloved SciPy would get lost underneath all the competing packages.
So he made a choice that would define his career. He had a tenure-track position where they wanted him to write papers. But Travis knew that writing NumPy would have a bigger impact. He gave up that secure path and spent his time unifying the community around a single, powerful array concept instead.
“I knew that more people would care about it” he said simply.
But NumPy didn’t succeed because of Travis alone. It succeeded because of the 694 contributors who came to the project and added their work, their knowledge, their passion. It succeeded through 144 releases, each one improved by people who cared.
At every critical juncture, Travis chose people over technology. He was up late at night writing documentation—not glamorous work, but work that helped others. He started businesses like Anaconda and Quansight not because he loved business, but to solve a human problem:
how do you pay open source developers?
Today, NumPy is everywhere. TensorFlow and PyTorch were built inspired by it. Financial institutions use it to understand risk and avoid crises. Physicists analyze data with it. Travis and a few hundred contributors literally altered the course of human events—not because they were geniuses, but because they focused on helping people and built a community around that shared purpose.
The Lesson That Changed My Perspective
Before, this I always thought Open-source contributions meant being smartest person and need to have a world chaning idea. He showed me that open source contribution is fundamentally about helping people. One person. Ten people. Eventually millions. But it starts with that simple desire to make someone else’s life a little easier.
He reminded us that the opposite of acceptance isn’t critique—it’s apathy. The worst thing that can happen isn’t that someone disagrees with your code. It’s that nobody cares at all.
And he revealed something profound: your brain can’t really distinguish between helping five people and helping a million people. The satisfaction comes from the act of service itself, not the scale.
This realization was liberating. I didn’t need to create the next NumPy to make a difference. I could:
- Fix a bug someone reported
- Write documentation for a feature I’d just learned
- Contribute an example that helped me understand something
- Answer a question on Stack Overflow
Each of these acts helps someone. Each one matters.
Watching Travis’s talk in 2018 changed something fundamental in how I think about my own work. It gave me permission to start small, to focus on service rather than recognition, and to remember that the people in my life—family, friends, collaborators—are more important than any code I’ll ever write.
It also gave me hope. Because if a graduate student working late at night while his family slept could create something that changed the world, then maybe I could make a difference too. Not by being extraordinary, but by being consistent, helpful, and focused on people first.
Building your own journey
Travis gave practical advice for getting started: search for projects that match your interests, look for issues tagged “help wanted” or “good first issue,” write tests, contribute examples. But underneath all that practical guidance was a deeper message.
Start with what you care about.Be passionate about something.Think about what makes you come alive, that’s where you should do some work.”
For him, it was medical imaging and signal processing. For you, it might be web development, data visualization, game engines, or something entirely different. The technology doesn’t matter as much as the people you’re helping.
He also reminded us that we’re going to need teams. No one builds something significant alone. And to build a team, people need to like being with you. Nobody wants to work with someone who’s grouchy, entitled, or focused only on being right.
If you’re a leader and you’re not thinking about serving others he said, “then leadership is beyond you.”
The NumPy story isn’t really about arrays and scientific computing. It’s about what happens when you combine technical skill with genuine care for others. It’s about building things that matter by building communities that care.
And that’s a story that inspires me every time I consider contributing to open source—or really, every time I sit down to write any code at all.
Why this talk mattered to me?
Watching Travis’s talk in 2018 changed something fundamental in how I think about my own work. It gave me permission to start small, to focus on service rather than recognition, and to remember that the people in my life—family, friends, collaborators—are more important than any code I’ll ever write.
It also gave me hope. Because if a graduate student working late at night while his family slept could create something that changed the world, then maybe I could make a difference too. Not by being extraordinary, but by being consistent, helpful, and focused on people first.
The NumPy story isn’t really about arrays and scientific computing. It’s about what happens when you combine technical skill with genuine care for others. It’s about building things that matter by building communities that care.
And that’s a story that inspires me every time I consider contributing to open source—or really, every time I sit down to write any code at all. Eventually I also could build a python package - whisper-normalizer with 1.6M+ downloads at the time of writing.
Your Turn
If you’re reading this and wondering whether you can make a difference in open source, Travis’s answer is clear: yes, you can. Start where you are. Help one person. Then help another. Build relationships. Serve your community.
The technology will follow. But people come first.