AI turned tech leads into architects

Jack Franklin, a senior software engineer at Google, says AI lets one engineer plan across the stack, so his tech lead job moved to keeping work consistent and reviewing plans.

Jack Franklin, a senior software engineer at Google, wrote in LeadDev about how his work as a tech lead has changed. An engineer with an AI model, the codebase and its history can now propose a design spanning front end, back end and database with little help. Coordination before work starts matters less, and review of the resulting design matters more.

• He describes his role as closer to an architect. He sets the shape of the system and its goals, and keeps the team consistent as each person moves faster.
• He sets limits for the AI agents. Examples include improving the instructions a team gives its agents and building command-line tools that give an agent accurate data.
• When an engineer uses AI to propose changes outside their area, the right domain expert must review the plan. AI-generated mock-ups do not replace a UX designer's real designs.
• He warns that AI will make questionable decisions to get tests passing. The result is subtle bugs, overly complex code and outages at 4am that nobody on the team understands.

The tech lead's review of design documents, and the choice of who else must read them, becomes the main point at which an engineering team sets quality. That teams should write this review into their process is an inference and is not the author's claim.

Derived by Working Surface from the article; more in the Playbook. Source line: AI isn’t an excuse to not involve humans.

4 November 2026: teams building with AI agents are turning each failure or correction into a permanent check, rule or shared record, rather than fixing one result at a time.