The threat models here aren't phishing kits or ransomware - they're orbital debris collisions, electromagnetic interference, and the fidelity of a laser weapon simulation under adversarial conditions. Stellar Science, founded in 1999, builds the software that lets government agencies and defense contractors actually trust their computational analysis. The domains read like a DARPA wish list: space domain awareness, computational electromagnetics, high-performance computing, and laser weapon system simulation. This isn't a consultancy billing hours for PowerPoint; it's a shop that delivers operational tools in domains where a modeling error has kinetic consequences.
The team is built on a specific hiring profile: programmer-scientists. Not scientists who can hack together a script, not developers who took an orbital mechanics elective - people with deep subject matter expertise who also write robust, maintainable production code. The distinction matters when the software is ingesting sensor data to track objects in low Earth orbit or running electromagnetic simulations that inform real procurement decisions. Computer vision pipelines for defense applications don't get the luxury of "we'll fix it in the next sprint."
Stellar Science operates at the intersection of computational science and mission-critical software engineering, bridging research and deployable systems for clients who can't afford to be wrong. The work is technically dense, the stakes are legible, and the tolerance for hand-waving is zero.
