SpaceXAI - Grok Model Evaluation
Completed software-engineering model evaluation contract, May to September 2026
Overview
From May to September 2026, I contracted with SpaceXAI (formerly xAI) on Grok model evaluation, reviewing AI-generated code and assessing how well it met real software-engineering requirements.
Work Scope
- Evaluated AI-generated code for correctness, maintainability, performance, security, and alignment with requirements.
- Reviewed software-engineering model outputs and wrote evidence-grounded assessments of their quality.
- Worked across multiple programming languages and domains, drawing on production engineering experience.
Why It Matters
The work applied production engineering judgment to coding-model quality: checking whether generated software satisfied the task and explaining why it worked or failed.