Cloud sandboxes to run your full stack (https://eng.somethingelse.ai/). Primary use case is with PMs to prototype and build on the codebase, shipping non-opinionated PRs for review, but our dev is using Else to build Else, so we published the eng site so others could try it out.
Personally, I'm working on a river TierBlend seasonal forecast (currently 6–15 weeks per gauge) — in-house ML model trained on 35+ years of weekly discharge plus NRCS SNOTEL snowpack features (https://pauldjohns.github.io/usgs-discharge-poc/)
Personally, I'm working on a river TierBlend seasonal forecast (currently 6–15 weeks per gauge) — in-house ML model trained on 35+ years of weekly discharge plus NRCS SNOTEL snowpack features (https://pauldjohns.github.io/usgs-discharge-poc/)