NumFOCUS

Congratulations to our SciPy 2026 Diamond Sponsor, @Nominal_io, on the launch of Instro, their new open source library for Python. If you're joining us at SciPy this week, stop by Nominal's Lunch & Learn to see Instro in action, meet the team, and learn more about what they're building. #SciPy2026 #OpenSource #Python

Watch Give Me 5 S2E8 at: youtu.be/qSSdZbkHH1s

Yujian Tang has a hot take on AI — and he's not pulling punches: "AI is nowhere near as good as Twitter wants you to think it is." Text models? Solid. Image generation? Getting there. Video generation? Still rough. Audio? Not quite. His point isn't that AI is overhyped full stop — it's that the hype is outrunning the actual capability, and social media (especially Twitter) is doing the most to widen that gap. "We're in such an early stage. People think AI is a bubble. I don't think we're even anywhere near what the peak is going to look like." The real frontier? Multimodal tools. Agentic systems. And open source communities that actually document how to use them. "It's still the people." From Give Me 5 Season 2. Full episode in comments.

Check out GM5 S2E8 here: youtu.be/qSSdZbkHH1s

Yujian Tang's advice to anyone new to software, data, or AI? Don't take a surface-level approach. "Start diving into how things work as quickly as you can. Once you get to a certain level — go in and just read the code." And when you're stuck: talk to people on forums. Go to live events. Ask questions. Open source makes this easier than ever — the tools are free, the communities are open, and the people who built them are often in the same Slack channels and meetups you are. From Give Me 5 Season 2. Full episode in the comments.

Yujian Tang started coding at 13 — not because someone told him to, but because he thought it was cool. That curiosity took him from NumPy and SciPy in college, to Amazon, to Zilliz (the team behind Milvus, one of the first open source vector databases), to working on some of the earliest LangChain and LlamaIndex builds — before most people knew what an LLM was. Now he's CEO of OSS4AI and a collaborator with PyData chapters across the US. "Open source gives you a very low cost of entry. It's basically free to use these tools — and it helps you learn a lot about the technology." Give Me 5 Season 2, Episode 8 drops tomorrow. 👀

OSS4AI founder Yujian Tang shares a striking data point from a recent San Francisco hackathon: out of 40 projects built in a single afternoon, more than half came from one-person teams — and most of the winners did too. In this clip from NumFOCUS's b.o.s.s. Town Hall, Yujian unpacks how coding agents are reshaping what's possible for solo builders, and why that's good news if you've been sitting on an idea. Watch the full panel: hubs.la/Q04nyp_s0

Daniel Brumund, an advisor on AI and digital governance at GIZ, has spent years supporting open source language technology projects across Africa and Asia. In this clip from the b.o.s.s. Town Hall, he explains why true open source AI requires showing up with humility, leaning into the "necessary friction" of community engagement, and asking who actually benefits from "open" data and tools. Watch the full conversation: hubs.la/Q04nxFS10

Noor Aftab, Chair of the NumFOCUS Board of Directors and a Global Program Lead at Amazon, shares startling research on representation in AI and open source: women make up just 22% of data and AI roles — and only 3% of open source commits. In this clip from the b.o.s.s. Town Hall, Noor explains why this isn't just a pipeline problem; it's a structural one, with real consequences for who AI systems serve well and who they fail. Watch the full conversation: hubs.la/Q04nwHnH0

@teoliphant, CEO of Open Teams, creator of NumPy, SciPy, and Numba, and co-founder of NumFOCUS, explains why he thinks AI's real test is simple: does it help people, or doesn't it? In this clip from NumFOCUS's b.o.s.s. Town Hall, Travis talks about the human connection that drew him to open source in the first place — and how AI is already changing community dynamics, including a rise in "lurkers" who get answers from AI instead of showing up in forums and mailing lists. Watch the full talk at hubs.la/Q04nqtls0

How Is AI Showing Up in Open Source Communities? We tackled that question during the b.o.s.s. (best of open source science) town hall on June 25, 2026. AI is moving fast, and open source communities are feeling it from every direction — contributors, maintainers, funders, and policymakers alike. In this Q2 edition of the b.o.s.s. town hall, NumFOCUS Executive Director Rachel Kerestes moderated a panel of four leaders working at the intersection of AI and open source to unpack what's changing, what's at stake, and what comes next. The panel: @teoliphant— CEO of Open Teams and co-founder of NumFOCUS @nooraftab — Global Program Lead at Amazon Web Services Daniel Brumund — Advisor for AI and Digital Governance at GIZ Yujian Tang — Founder of OSS4AI Watch the full session now on the NumFOCUS YouTube channel at: hubs.la/Q04nhJCh0

A few hours left.   Fuel the future one last time this summer. Every dollar goes directly to keeping the open-source tools powering modern science alive, maintained, and free.   Max the Stack: hubs.la/Q04n6r-f0

Today is the day.   Our summer campaign closes at midnight. If you haven’t given yet, open science needs you to answer that call today. Max the Stack today at hubs.la/Q04n6mXN0

You wouldn't let your dev environment go unmaintained all summer. Don't let the open-source tools powering real science go unfunded either. Back the stack before June 30. hubs.la/Q04m-37Q0

The days are long. The deadline is short. Our summer campaign closes June 30 — and open science needs you before then. Every dollar you give goes directly to sustaining the tools keeping modern research alive through summer and beyond. Max the Stack today at hubs.la/Q04mV49T0

Open-source science doesn’t take a summer break.   Artemis II astronauts captured a solar eclipse from the Moon this spring — and SunPy and AstroPy were there, used to process and analyze those historic images. This is what funded open source looks like.   The most important scientific tools are invisible. We make sure they keep running. Back the stack this summer. Max the Stack today at hubs.la/Q04mHrV30

TOMORROW: Join us for our Q2 b.o.s.s. town hall tomorrow! We’re bringing panelists from across the open source scientific community to talk about the question “How A.I. is showing up in O.S. Communities?” Share this with your communities, we’d love to have a broad group of projects and contributors join the conversation. Thursday, June 25, 2026 9AM PT | 11 AM CST | 12 PM EST | 6 PM CET Meet the panelists: Travis Oliphant | CEO/Founder, OpenTeams | Creator of NumPy, SciPy, Numba | Co-founder, NumFOCUS Noor Aftab | Global Program Lead, Amazon | Chair, NumFOCUS Board | IBM Champion ’25 | TEDx Speaker Daniel Brumund | Advisor: AI & Digital Governance, GIZ | Former FAIR Forward | Equitable Data Governance Yujian Tang | CEO, OSS4AI | Prev. Stanford, Zilliz, Amazon | Educator, Investor, Builder Join the Zoom Webinar: hubs.la/Q04mylQB0

Meet one of the panelists for the next session of the b.o.s.s. (best of open source software) town hall series! This session centers on the topic: How A.I. is showing up in O.S. communities. Yujian Tang, CEO, OSS4AI | Prev. Stanford, Zilliz, Amazon | Educator, Investor, Builder Yujian Tang is the founder of OSS4AI, one of the largest open source developer communities in the world. He has been a guest lecturer at Stanford, a software engineer at Amazon, and a published researcher. He brings a rare combination of community-building, commercial, and investment experience to the question of how AI is reshaping open source contribution and sustainability. Join the conversation by registering online at hubs.la/Q04mjsJn0

Elizabeth Osanyinro is building PyData Bradford between Leeds and Manchester — and she's got one clear ask for NumFOCUS: A speaker bank. "When you're short of speakers, it can be really difficult having to send cold emails to potential speakers and all of that. If we had a speaker bank — oh, I need a speaker! NumFOCUS, can you get me one?" It's a simple idea with a massive potential impact for local PyData chapters everywhere — especially in smaller cities where networks are thinner, and outreach is harder. MK noted it live: first time they've heard this one. And it landed. I f you're someone who speaks on data, AI, or open source topics and wants to get connected with local communities, drop a comment. Watch the full episode here: hubs.la/Q04mjLR-0

Meet one of the panelists for the next session of the b.o.s.s. (best of open source software) town hall series! This session centers on the topic: How A.I. is showing up in O.S. communities. Daniel Brumundm, Advisor: AI & Digital Governance, GIZ | Former FAIR Forward | Equitable Data Governance Daniel Brumund is an Advisor for AI and Digital Governance at GIZ’s Competence Center “Digital Societies,” focusing on AI governance, regulation, ethics, open-source licensing, and NLP. From 2022–2025, he worked with FAIR Forward, GIZ’s initiative promoting open-source AI for local innovation across Ghana, Kenya, Rwanda, South Africa, Uganda, Indonesia, and India, where he led work on community-driven data collection, AI in low-resource languages, and equitable data governance. He brings over 15 years of experience in international development at the intersection of digitalisation. Join the conversation by registering online at hubs.la/Q04m3YFK0