Sean Jungbluth is a Claude Community ambassador and the founder of BioKEA, a Berkeley lab working on biodiversity and environmental DNA. BioKEA is building an open-source environmental workcell to help turn field samples into biological knowledge: preparing specimens, imaging and sorting insects, and moving samples into plates for DNA barcoding.
That work connects directly to our mission at Worldwide Studios: giving scientists hands-on ways to build with the latest AI and hardware. Our partnership with Sean and Claude Community brings that shared purpose into a series of gatherings where researchers can share tools, question assumptions, and work on problems from their own labs.
24hr AI for Science World Models Hack @ Zeon Systems
We first met Sean when he joined us for his first hack at Zeon Systems. With Jeremy Garibay and Austin Baker, he integrated an xArm into BioKEA’s BugPicker and imaging workcell, connecting specimen handling with the movement of plates between instruments.
Their open-source plate-shuttle workflow moves a 96-well plate from the BugPicker to an imaging station and back. It is one practical piece of BioKEA’s broader environmental research pipeline, and a starting point for other scientists to build on. That first weekend together led to the three events below.
Across the hack, eight teams shipped projects in 24 hours. The work ranged from closed-loop biology to robotic dexterity, with a recurring lesson: generating a protocol is only part of the job; teams also need to verify that the physical action actually happened.
Claude Community SF: Science Demo Night
At Humanmade, nine demos filled three hours with working tools built with Claude Code. Austin Baker showed BioKEA’s computer-vision pipeline for a vacuum-based insect sorter that moves specimens into well plates, while Brontë Kolar demonstrated Zeon Systems’ world-model platform for composing robotic workflows and previewing them in simulation.
Simon Coelho shared HiveMind, an open cell-culture monitor designed for Openshelf. Rapid demos widened the conversation to experimental-data copilots, protein-function model training, robotic peeling, and agentic fly tracking—giving researchers concrete workflows to compare and people to follow up with.
View Science Demo Night on Luma · Read the recap
SF Claude Conversation: How do Credentials Shift in Science?
Our September 29 dinner salon at Hanwha AI Center, “How do Credentials Shift in Science?”, explored the limits of traditional training as robotics and AI enter the lab. The questions were concrete: what does a PhD establish as a credential, what does a good human–AI loop look like, and how does a scientist’s role change when experiments run more autonomously?
The gathering was designed for scientists, researchers, and technical builders using Claude Code for automation and computational work, with curious newcomers welcome. Dinner and an open discussion gave people room to compare the skills, knowledge, and networks needed to keep up with new tools.
View the science salon on Luma · Read the original invitation
SF Physical AI for Science Impact Lab
On October 17 at Studio 45, the Physical AI for Science Impact Lab brought scientists, software engineers, and hardware builders together for a full day of hands-on exploration. Teams worked on sample preparation, instrument integration, and open automation challenges, with time to build and share demos of new benchtop workflows.
The day moved from a morning kickoff and team formation into an extended build session, then evening demos. BioKEA provided the environmental lab context: how to make the steps between a field sample and a usable biological result more reliable. With no prerequisites, the format made room for people to bring different strengths to the same physical problem.
View the Physical AI for Science Impact Lab on Luma
A shared mission
Worldwide Studios and Claude Community share a mission: empower scientists and researchers to discuss, debate, and build with the latest tools to create new solutions in the lab. We want more people to have the space, community, and hands-on experience to turn a question from their own research into something they can test.