THE INFORMED · SINCE 2020
A small club of misfit scientists
Seven scientists and service members who find problems in Navy meteorology, oceanography, and space that nobody funded, build the fix, and hand it over free.
What we do
We're all in the military, so we can't be paid by it, and we don't want to be. We've never taken a dollar or asked for one. H BAR LLC exists so the group has a legal entity; it doesn't collect anything.
The pattern is the same every time. Someone in the community has a problem nobody has funded. We build it, show it working, and hand it over. Then the command knows what to ask the Office of Naval Research for, and money moves to the labs to build the real thing.
Some of what that has looked like:
A weather and ocean viewer the Navy could own
The Navy had tried for years to replace its reliance on a commercial site. We built one that works at low bandwidth, afloat, and then went further: trained models that nowcast radar returns sixty minutes out, reconstruct radar from satellite where there is no ground radar, super-resolve imagery, and rebuild true color. The first weights finished in September 2026 and training continues. They did their job as a demonstration: the command asked what we had, then whether we had really done it on two consumer graphics cards, and now has proven capability to take to the admiral when asking the Office of Naval Research for funding.
Teaching tools Sailors get before the real thing
Information Warfare Sailors often don't touch Builder, TAWS, or PC-IMAT until they deploy, and concepts like the electromagnetic spectrum are hard to teach without something to click on. So we built the "Not" sites (Not Builder, Not TAWS, Not PC-IMAT), the Skew-T Explorer, the EM Spectrum Explorer, and the Cryptography Explorer.
Space
Maritime Space Officers needed to see what space weather does to orbits and links. The Space Operations Explorer showed it live. They asked for more, so OrbitWatch went through the public catalog to find orbit changes, catalog changes over time, and patterns of life. Two preprints came out of it.
Advocacy through research
We lost microwave sounders and scatterometers over the years, and a post-mortem said the reason was advocacy, not science. The BAMS paper on microwave data in tropical cyclone intensity estimation was written to fix that, alongside a classified report. It helped win increased remote-sensing support; last year the Pacific got its first airborne radar mission into a typhoon.
AI literacy for the department
The GenAI Field Guide teaches people how to use generative AI correctly at work. It runs on AWS GovCloud behind a Common Access Card; an open edition without the DoD specifics is public.
Connecting the lab to the fleet
The Navy has excellent scientists at NRL and NAVO, and commanders with hard operational questions. We spent three weeks at NRL Monterey helping each side hear what the other needed. GraphCast work continues there as a separate effort.
Who
I founded it, and seven of us work on it every day. The four of us below lead it and work with the agencies. We're equals; every partner builds.
- Taylor Wulff-Morrison, MAJ · agentic AI.
- Derek Conklin, AG1 · operational impacts. He deploys more than any of us and runs the rack-and-stack where we decide what to take on.
- David Robinson, LT · electromagnetic spectrum. The EO/IR sites are his and Taylor's.
- Sean Egan, LT · earth and space science, machine learning, and keeping the whole thing running.
Where it started
I was a typhoon forecaster, then the operations officer, at the Joint Typhoon Warning Center in Pearl Harbor. It's a technical command; the tools forecasters use are written and maintained by whoever at the command can code. That was three people. Meanwhile there were Sailors and Marines walking around with real programming backgrounds that nobody had thought to use.
So I founded the JTWC Development and Innovation Group and got it written into command directive: a formal coding group open to Sailors, Marines, and civilians, with a training pipeline, regular seminars, code review, and Scrum. Three developers became twenty. Projects that had sat shelved for years shipped within months. The commanding officer and her admiral liked it enough that sister commands copied it.
The Informed came out of the same time. Beyond coders, I kept finding people who loved the science and wanted to use it to help the commands under Naval Meteorology and Oceanography Command. We decided to be a group that looks at a weather, ocean, or process problem, builds the answer, and gives it away.
Why free
Because it's fun, because we're scientists, and because we care about the people in this community. The DoD supercomputing behind this is granted to me as a researcher; the hours are my own. When I left active duty I picked up the pace because I wanted to learn how far agents and models could go. They went further than I expected.