Delivery
● Lead sprint delivery for a team of 6–12 engineers: planning, dependencies, risks, release readiness.
● Think about how work gets sequenced across people and AI-assisted workflows, not just story points.
● Keep JIRA, ProductBoard, and the dependency tracker current. No stale tickets, no surprises at SoS.
● Track cycle time and escaped defects; own the trend, not just the number.
Technical
● Review code and AI-generated output for whether it solves the right problem. Clean tests aren’t enough.
● Define what needs human sign-off, what can be spot-checked, and what warrants a second pass.
● Write specs that work as contracts: clear scope, explicit constraints, stated boundaries.
● Stay close to architecture decisions, tech debt, and incident response.
● Contribute to software architecture decisions and patent writing for innovative work.
People
● Run 1:1s that are about growth, not sprint status.
● Coach engineers on when to trust AI output and when to push back. Model it explicitly, not just in retrospect.
● Give clear, timely feedback; build real development plans; address performance early.
● Help your team see that directing AI well is a skill worth building, not just a tool to adopt.
● Calibrate hiring panels, provide written feedback, contribute to calibration discussions.
Cross-functional & Culture
● Be the PM partner who holds a definition-of-ready and surfaces risks before they land in a leadership meeting.
● Engage escalations directly; produce RCAs; own P1/P2 support SLAs.
● Set the tone on your team: accountability, directness, recognition of others’ work.
● Watch for workload and wellbeing signals before they become problems.
● Lead by influence across teams and functions ; comfortable driving cross-team technical decisions without direct authority