Manager, Software Engineering

Role Overview

Lead the AI Platform Team at dbt Labs and build the Agentic Platform that powers internal AI features and external AI integrations.

Responsibilities

• Build, lead, and coach a team of 5–8 engineers focused on a robust, scalable AI agent platform.
• Create an “Agent-First” experience by building the APIs and services that enable agents to reason, plan, and execute within the dbt ecosystem.
• Define dbt’s MCP strategy and develop dbt MCP tools so LLMs (e.g., Claude, Codex) can:
  • fetch relevant context,
  • validate SQL,
  • understand metrics in their development environment.
• Bridge platform and product by partnering with product teams to deliver primitives such as memory, tool-calling, and reasoning loops for dbt agentic use cases.
• Coach engineers to produce “agent-ready” codebases, including deterministic outputs and optimized tool-use.
• Drive technical excellence by establishing standards for how agents interact with dbt Cloud, ensuring security, governance, and auditability.

Requirements (Mandatory)

• 3+ years leading high-performing software engineering teams (people management).
• Experience with agentic architectures, including an agent loop lifecycle (Plan → Act → Observe).
• Strong API design expertise; familiarity with standards like MCP or OpenAI Function Calling.
• Solid software engineering fundamentals, with a strong POV on maintaining dbt’s analytics engineering rigor (testing, CI/CD) in an AI-driven world.

Nice to Have

• Direct experience building or contributing to MCP servers.
• Experience with the Modern Data Stack, including the dbt Semantic Layer as a “source of truth” for AI.
• Excellent written communication for a remote-first culture (documentation and rule-of-engagement clarity).

Additional Signals

• Interest/passion for developer tools and understanding how tools like Claude Code / Codex are changing data engineering workflows.
• Experience with orchestration/state & context management for LLMs.
• Ability to collaborate with external AI labs and internal teams to make dbt the preferred “data context” layer for major LLMs.