PraxisAI

0 reviews

PraxisAI helps hiring teams assess software engineers on how they actually build with AI. Candidates fix real repository bugs (SWE-bench issues from projects like Django and Flask) or complete 24-hour take-home projects in a browser VS Code workspace with a coding agent, a timer and a token budget. Grading tests check the outcome, and recorded prompts, agent actions and test runs give reviewers evidence for a process review across six scored dimensions. Teams can set token budgets per assessment. A candidate-side practice product (praxisai.site) lets engineers rehearse the same format.