Analytics Engineer

Role Overview

As an Analytics Engineer at Addi, you will own the marketing data platform, map the data pipelines and integrations, and ensure reliable, efficient, and self-service data for marketing teams. You will work on a complex SQL platform, build clean and reusable data tables, and automate data work with AI agents.

What You Will Do

Your main responsibilities will be to ramp up the marketing data platform in 90 days, add tests and alerts to critical pipelines, and ship AI agents or agentic workflows into production. You will also reduce the cost or runtime of main marketing pipelines by at least 30% and maintain a catalog of marketing data with owners and approval rules.

Why It Might Be a Fit

This role is ideal for someone who has proven experience in writing and debugging complex SQL, building ELT pipelines, and automating data work with AI agents. You will have the opportunity to work on a problem that truly matters, be part of something big from the ground up, and enjoy unparalleled growth opportunities.

Requirements

• Proven experience writing and debugging complex SQL without AI
• Writes and reads joins at scale, window functions, CTEs, and incremental logic with confidence
• Can tell when a query result is wrong and explain why
• Builds clean, reusable tables and datamarts that are easy to extend, test, and document
• Track record of building and operating ELT pipelines on a modern data platform
• Hands-on with dbt, Airflow or similar, plus Git and code reviews
• Has used SQL warehouses, notebooks, and scheduled jobs, and understands catalog, schema, table, and permissions
• Uses tests, monitoring, and alerts by default; comfortable with on-call and fixing issues at the root
• Demonstrates ability to automate and integrate with Python
• Builds scripts and API integrations that others can run without the author
• Experienced in using coding agents daily for real data work
• Uses Claude Code, Codex, Cursor or similar, not just autocomplete
• Writes a spec before building (inputs, outputs, transformations, checks), delegates multi-step tasks to the agent, and reviews the result
• Can show a before and after
• Has solid expertise in building agents and agentic workflows that others use
• Has built at least one agent or workflow that chains LLM calls with tools (SQL, APIs, MCP servers) and that other people actually use
• Understands failure modes, guardrails, and how to test an agent with real inputs
• Demonstrated background in validating AI output
• Has caught AI work that looked right but was wrong (e.g., a plausible SQL query with a bad join) and added a check so it doesn’t happen again
• Knows when human review is required, especially for customer data sent to marketing tools
• Track record of writing documentation that people and agents can run on
• Writes context and runbooks (READMEs, agent instructions, data catalog entries) that let a teammate or an AI agent maintain the system without the author
• Proven ability to translate business needs and own outcomes in ambiguity
• Understands what Marketing, Growth, and Marketplace need and decides whether the answer is a table, a pipeline, an agent, or a self-service tool, instead of a one-off report
• Manages their own backlog and ships without close direction
• Communicates clearly with technical and non-technical people; English is a must

Benefits

• Competitive compensation
• Meaningful ownership
• Benefits that go beyond the basics to support growth