Engineering Manager

Responsibilities:

• Delivery, technical direction, and operational health for your team.
• Production services at scale, with the availability and correctness bar that high-traffic, customer-facing systems demand; AI agents that act on a customer's behalf, with the guardrails, idempotency, and audit trail that demand.
• Conversational quality across chat and voice, including real-time and multilingual interaction.
• Outcomes, not output: you measure the business impact of what ships and build the evals that catch regressions before customers do.

Requirements:

• 8+ years in engineering, including 2+ years managing engineers directly.
• Depth in backend fundamentals: distributed systems, API and event-driven design, data modelling, idempotency, caching, observability.
• You've owned high-traffic services and the incidents that came with them.
• LLM systems shipped and operated in production: retrieval, tool calling, orchestration, guardrails, evals, fallback behaviour, latency and cost budgets.
• Hands-on: you prototype, review designs and code, and can build an agent or eval harness yourself when that's faster than explaining it. Not carrying sprint work, but never far from it.
• Daily fluency with agentic coding tools, and a clear view of where they help and where they create review debt. Judgment about when an agent should act on its own, when it should confirm, and when it must hand off to a human.
• How you lead: You grow engineers deliberately, and people who have worked for you want to again. You hire well and raise the bar with each hire. You hold a technical position in a room of senior engineers without holding the keyboard. You're steady when scope shifts, and you build teams that outlast any particular charter.
• Strong signals: Deployed agentic products on customer-facing surfaces: agents that took real actions for real users at scale, with the blast radius that implies. Voice agents are a strong plus.
• Multilingual or code-mixed conversational design; high-fanout delivery or messaging systems.
• Range across more than one language and runtime, with the ability to review design and code in an unfamiliar stack.
• Built an eval harness or internal agent framework that other teams depended on.
• Replaced a rules-driven flow with a model-driven one and can explain what broke.