element451 - Team Lead, Delivery Management

• 6+ years of professional software engineering in a complex, multi-tenant SaaS product, including meaningful hands-on technical work. • 1+ years leading engineers — as a manager, tech lead, or team lead. This can be your first formal management role, but you’ve owned outcomes through other people before. • Enough technical depth in our stack — PHP/Laravel, TypeScript (NestJS/Angular), MongoDB, AWS — to engage substantively in architecture discussions, review PRs, and make sound trade-off calls. • Hands-on experience delivering customer-facing AI or agentic capabilities in production, with working knowledge of knowledge grounding and retrieval, tool use, evaluation and regression testing, human-in-the-loop design, observability, latency, reliability, safety, and cost. You can speak concretely about how you evaluated behavior, managed failure modes, observed production performance, and released changes safely. • A track record of delivering — owning velocity, predictability, and quality for a team or significant scope, all the way through to production. • Comfort operating within a cycle-based product development methodology; familiarity with Shape Up or a comparable approach is a plus. • A genuine commitment to building an AI-forward team. We’re a Cursor shop; Claude Code, Codex, or comparable tools are equally valued. • Strong communication at both the engineering-detail and leadership-summary levels — effective in 1:1s, team retrospectives, product planning, incident reviews, and leadership reviews alike. • Development Lifecycle • This appendix defines phase-by-phase expectations for how a Team Lead operates across the development lifecycle for a single team. The Director owns the system that makes this work across teams; the Principal Engineers own technical standards and architecture. • Owns cycle shaping for the team — works with Product, the team’s engineers, and Principals to size and scope work to a deliverable cycle boundary, grounded in honest analysis of capacity, complexity, risk, and cross-team dependencies. Ensures the team participates actively in refinement and represents the team’s capacity to Product before commitment. • For AI work, ensures that intended behavior, success measures, known failure modes, data or evaluation needs, and rollout expectations are explicit enough to support an engineering commitment. • Ensures a design process is applied when a feature has broad implications, material AI behavior risk, or significant architectural risk, partnering with the Principal Engineers who own the design process. Makes sure design outcomes are documented and that raising the need for a design is treated as sound judgment, never a delay. • Ensures design considers the full product path: knowledge and data dependencies, agent tools and permissions, human escalation, shared platform boundaries, observability, evaluation, latency and cost constraints, and failure behavior. • Holds the team to consistent engineering standards — code quality, PR hygiene, review thoroughness, and unit testing — reinforcing the standards the Principals own. • Monitors progress without micromanaging: aware of where each engineer is, surfaces blockers early, and removes obstacles. • Enforces the team’s testing obligations as part of done — unit tests ship with the code and integration tests cover paths that cross boundaries. Owns the team’s engineering–QA contract and resolves friction at the engineering–QA boundary before it becomes a delivery problem. • Ensures AI behavior is tested with methods appropriate to the risk: deterministic checks where possible, representative regression datasets, calibrated evaluation where judgment is required, adversarial and failure-path coverage, and human review for sensitive workflows. • Owns release health for the team — “done” means in production and operating as intended. Drives completed work to release, follows the org-wide release standards the Director owns, and removes blockers that delay deployment. • For material AI behavior changes, ensures rollout and rollback are deliberate, production signals are monitored, and regressions or unexpected cost, latency, safety, or quality changes have a clear response owner.