Mission engineering for small-satellite teams
One spacecraft model. Every result hangs off it.
Keep the whole study on one model, from the first brief to design review. Runs in the browser.
Runs in the browser. Pilot access is currently by request.
Surface Studio · 6U assembly
FAQ
Questions that come up in review.
These are the questions engineers and review boards ask about scope, AI, and validation.
Need to ask about a specific mission? Ask us directly.
What does Paraxis actually do?
Paraxis carries a small-satellite mission from concept to design review. The mission, the spacecraft, the analyses and the review evidence all read one model, so the study stays consistent as it changes. It runs in the browser.
Where does the AI fit in?
AI drafts mission configurations, simulation scripts, and report text. An engineer reviews and approves that work. Versioned physics engines calculate the numbers, and verification checks the design again after a change. AI does not set an engineering verdict.
Which AI models does it use?
Paraxis can use commercial models from providers such as OpenAI and Anthropic through their APIs. The provider is a configuration choice. Physics, stored mission data, and the engineering workflow do not depend on an AI vendor, and the platform works with the assistant switched off.
Is my mission data used to train AI?
No. Assistant requests send only the context needed for that one draft, and nothing you store in Paraxis is used to train any model. AI features are optional and inactive until you invoke them.
How do I know the physics is right?
Maturity differs by domain. Link budgets reproduce NIST and US GSA reference calculations exactly. Three published flight-programme budgets reproduce within 0.09 dB for SwissCube, 0.05 dB for ESTCube-1, and 0.39 dB for HYPSO-2. The solar-array model matches MinXSS-1's published capability to 0.01%, and orbit geometry cross-checks against Orekit 21 for 21. The campaign logged 29 defects, closed 25 with regression tests, and retracted four early findings. Thermal and radiation currently use standard textbook models while their validation campaigns are built. Analysis pages label fidelity and provenance so modeled estimates are not presented as flight heritage.
Does it replace STK or GMAT?
No. Paraxis covers the work before those tools. It keeps the early design consistent and ready for review. Specialist analysis stays in specialist software, which Paraxis also uses for cross-checks.
Who is it for, and how do I get access?
Paraxis is for university satellite teams and small mission teams that currently coordinate work through spreadsheets or desktop tools. It runs in the browser with nothing to install. Access is by request while structured pilots are running.
Put the failed margins
on the table early.
Paraxis is available through structured pilot access for university and small mission teams.