Turn a bottle into a lab.
Start with what's in your recycling bin. Add a camera. Connect free AI analysis tools. Run a real experiment that generates real data: no grant, no specialized equipment required. Built on 30+ years of NSF-funded Bottle Biology research, decades of classroom teaching, statewide citizen science, and research-instrument entrepreneurship.
"I co-founded a company that built amazing research instruments. At $5,000 to $20,000 they were beautiful, but most people could never use them. They could not scale in college and university research labs, and certainly not in classrooms. SciBuilder is what I wish we had built first: real science that starts under $30."
Travis Tangen, Founder
SparkBottle (recyclable bottles become experiment chambers), SparkPanel (flat panels become observation walls), SparkBox (boxes become enclosed labs), SparkEye (the camera and timelapse rig), SparkGlow (the programmable LED light rig), SparkLab AI (free tools that make sense of your data), and the SciBuilder Token (the pointer that keeps your data connected): connected through universal SparkBlocks.
SciBuilder is in active development and is being co-designed with researchers, teachers, students, and the DIY at-home community during this phase. Three free SparkLab tools are live today. Physical kits are in development. Every label on this page tells you which is which.
A research lab works on its own data. A college course runs its own exercise. A high school class does its own activity. Even when everyone asks similar questions, the work rarely connects and the data rarely compares. The hard part was never curiosity. It was the activation energy: the cost, the equipment, and the coordination it takes before people can work together at all. SciBuilder lowers that activation energy, then builds the connections on top of it. Here is the order of operations.
SciBuilder deliberately takes on technical risk to push the scientific enterprise forward. The open question is whether accessible materials and a pointer token can really deliver research-grade, connected, trustworthy data at scale. That is what the early work tests, with clear checkpoints where we learn from what the materials and the token actually do, listen to the people using them, and adjust before moving forward. Collaboration is opt in, so groups can weigh their own participation and risk as they join.
Real science has long been built around the individual lab, with deep tradition, hierarchy, and expectation behind it. Asking science to be collective, simultaneous, and shared is a genuine departure. We think it is worth it, because the potential is just as large: far more people doing real science, together, earlier, with new and existing data brought into one living, shared inquiry.
SciBuilder is not just the bottles, the accessories, and the tokens. It is the order of operations that turns ordinary materials into a connected way of doing science.
Between a $3 toy and a $5,000 research microscope lies a vast, unoccupied middle. Educators, students, and curious minds are stuck choosing between playthings and instruments they can't afford, access, or understand.
The materials have been there all along: in your recycling bin, at the dollar store, in every classroom. What's been missing is the system that transforms them into real science.
The insight: bottles, boxes, and panels aren't just containers. They're science chambers. All they needed was a universal connector ecosystem, a family of instruments, and an AI partner to make sense of what you observe.
Try it yourself: Bug DJ, our organism path tracking tool, runs free in your browser right now. Open Bug DJ →
SciBuilder isn't a single product. It's a platform. Bottles, panels, and boxes all connect through universal SparkBlocks, turning everyday materials into a coordinated observation system.
One connector ecosystem. Three form factors. Infinite experiments. Any material. Any organism. Any question. Built from what's already around you.
No lab required. No specialized training. Just curiosity, materials, and a system designed to make real science accessible.
We lead with what's real. Three free tools are live right now, free starter print files are rolling out through the newsletter, and the first physical kit is in development. Everything else lives on the roadmap below.
Real analysis software, built first so you can do real science before a single kit ships. Open one in your browser and try it on your own footage today.
Free now. Free later. The tools stay free; kits and guides fund the platform.
SciBuilder's digital products arrive before the physical kits do. File packs, build guides, and video setup guides will sell as beehiiv digital products through the newsletter, and the free starter STL pack plus quickstart guide leads the way. Join the list and each release comes to your inbox as it lands.
Join the newsletter and the first free release (starter SparkBlocks print files plus a quickstart guide) comes to your inbox as it rolls out.
SciBuilder is designed for everyone who's ever wanted to observe, understand, and document the living world around them.
Nine domains, each with established protocols, curriculum connections, and AI analysis pathways built in from the start.
AI doesn't do the science. You do. AI helps you see what you built, tracking what the eye misses and measuring what the hand can't hold.
The platform collects huge, fast streams of observation data, and AI tracks it and makes sense of it. AI also helps us build the tools themselves. Different layers, different purposes.
"AI doesn't do the science. You do. AI helps you see what you built."
Here's what makes SciBuilder different from every AI toy on the shelf: you generate the data. The creativity to design an observation, the intuition to choose what to watch, the patience to document phenomena no one else has noticed: that's unlimited, and no algorithm can replace it.
But when your unique, in-person observations flow into AI analysis (growth patterns compared across hundreds of similar setups, behavioral data matched against reference frameworks, your single experiment joining a network of experiments asking the same question), the combination becomes extraordinarily powerful.
This work is moving through real channels: active solicitations and submissions with federal agencies including NSF, and ongoing conversations with industry partners. SciBuilder is actively pursuing NSF and USDA broader-impacts collaborations to bring these AI tools, along with sustainable reuse of materials, into bigger grant programs.
Human creativity designs the observation. AI finds the patterns you couldn't see alone. That's not automation. That's amplification.
SciBuilder isn't a startup idea. It's the synthesis of three decades of ecosystem observation research, hands-on science education, and deep partnership with the communities this platform serves.
SciBuilder is being developed in partnership with institutions that have been doing accessible science for decades.
Interested in being a pilot site? We're actively looking for K-12 classrooms, university programs, and informal science centers to partner with for the first wave of SciBuilder deployments.
Contact us about partnership →or email [email protected]
One recycled bottle. One free pack of SparkBlocks starter print files. One quickstart guide for your first build. Join the newsletter and the first free release comes to your inbox as it rolls out, plus early protocols, pilot invitations, and first-in-line kit ordering.
Just SciBuilder news from our home base in Madison, Wisconsin.