Writing code is no longer the top thing Augment Code looks for when hiring engineers. The AI coding tool company published a detailed framework this week that replaces traditional coding ability with six new evaluation dimensions, and the shift says a lot about where software engineering hiring is headed.
Augment's framework, authored by four members of its engineering and people leadership team, boils down to a simple argument: when AI agents can write most of the code, the engineer's job becomes deciding what to build, how to architect it, and how well they can direct AI tools to get there.
The Six Dimensions
The new criteria, in order:
- Product and outcome taste - Can you identify the right thing to build, not just build the thing you're told to?
- System and architectural judgment - Will your design survive production, or will it collapse under real load?
- Agent leverage - Can you actually turn AI coding tools into real engineering output, or do you spend more time fixing AI mistakes than you save?
- Communication and collaboration - Can you clarify ambiguous problems and build shared understanding across a team?
- Ownership and leadership - Do you drive outcomes end-to-end, or wait for task assignments?
- Learning velocity - How fast do you adapt as the tools themselves change?
The company also defined four distinct hiring profiles: AI-Native Systems Engineer (infrastructure focus), AI-Native Product Engineer (user empathy focus), AI-Native Applied AI Engineer (model understanding), and AI-Native Early Professional (optimized for learning speed over experience).
What Actually Changed
The before-and-after comparison is blunt. Traditional hiring valued writing code, implementing solutions, and individual output. Augment's AI-native model values specifying intent, orchestrating agents, choosing the right problems, and system-level outcomes.
This is one company's framework, not an industry standard. But Augment is a company that builds AI coding tools (they compete with Cursor, GitHub Copilot, and others), so they have a financial incentive to push this narrative. That said, the underlying logic is hard to argue with: if AI handles more of the typing, the humans who direct it need different skills than the humans who used to do the typing themselves.
The practical question for engineers reading this: how do you demonstrate "agent leverage" in an interview? Augment mentions evaluating candidates on their ability to direct and validate AI-generated work, but the blog post stays at the framework level without detailing specific interview exercises or take-home formats. That gap between theory and implementation is where most companies will struggle as they try to adapt their own hiring processes.