Lead AI Engineer for Production-Grade Customer Agents

Klaviyo; Boston (MA); Hybrid; USD 176,000 - 264,000; Full time

Job summary

Klaviyo in Boston, MA is seeking a Lead Engineer for Applied AI (Customer Agent). You will set the technical and product direction for a core area of Customer Agent, building a 0-to-1 agentic product with startup speed but the scale of an established tech company.

You’ll stay hands-on with the hardest problems, from orchestration to governance, and mentor others while shipping production-grade systems that handle real traffic.

Qualifications

• Built, scaled, and supported production agentic systems; ownership through launch, incidents, and post-demo operations.
• Experience securing agents in production against misuse and maintaining reliability as usage grows.
• Led technical and product direction for a significant area; shipped features others built on.
• Hands-on: writing code, reviewing architecture, debugging production issues.
• Patterns and designs reused by other engineers.
• Communicate complex trade-offs clearly across engineering and product teams.

Responsibilities

• Set the technical and product direction for the Customer Agent area from orchestration to governance.
• Tackle hardest technical problems: orchestration, evaluation, guardrails, memory and context, non-deterministic failure handling.
• Write and ship code, review designs, and make hard calls when there’s no obvious right answer.
• Define patterns that engineers adopt and mentor through design reviews.
• Collaborate with product, design, data science, and go-to-market leaders to shape what gets built.

Skills

Production agentic systems

Orchestration

AI observability

Incident response

Mentoring through reviews

Job description

Klaviyo in Boston, MA is seeking a Lead Engineer for Applied AI (Customer Agent). You will set the technical and product direction for a core area of Customer Agent, building a 0-to-1 agentic product with startup speed but the scale of an established tech company.

You’ll stay hands-on with the hardest problems, from orchestration to governance, and mentor others while shipping production-grade systems that handle real traffic.