Analytics Engineer

Location: New York City

Level: Mid-level, 3 to 5 years experience

As an Analytics Engineer at Moxxi, you will build and own the data models behind our marketplace reporting, and you will use them to answer questions the business has not yet figured out how to ask. This is a small analytics team, so you will not be handed a queue of tickets. You will work directly with account managers, yield, and product to understand what is actually happening in the business, then build the models and reporting that make those answers repeatable.

A significant part of this role is applying LLMs to analytics work: connecting models to the warehouse, engineering the context they need to be useful, and building the evals that tell you whether the output can be trusted. Our Analytics Lead owns this area, and you will be a key contributor to it.

Moxxi owns its ad server, CMS, user journey orchestration, and SMS platform, so we generate our own data and can change the systems that produce it. That also means analytics here covers both sides of the business. You will report on marketplace performance, and you will measure what product ships: how a change to a user journey affects conversion, whether a new feature did what we expected, and what that tells us to build next. You will be joining shortly after a platform migration, so expect to help define and instrument the events behind that work rather than inherit a mature measurement stack. Over time, we expect this role to grow into quantitative work on how we model and select offers.

• Own marketplace reporting across publishers, advertisers, and verticals, and make sure the numbers the business runs on are correct.
• Build and maintain dbt models in the warehouse, with tests, in version control, shipped to production.
• Investigate anomalies end to end: what moved, why it moved, and what we should do about it.
• Apply LLMs to analytics workflows, including context engineering and evaluation, in partnership with the Analytics Lead.
• Measure product work: define success metrics for new features and journey changes, specify the events needed to track them, and report back on whether the change did what we expected.
• Partner with operators to define KPIs so the business argues about decisions rather than about whose number is right.
• Build self-service reporting so routine questions stop routing through the analytics team.
• Own data quality: reconciliation across sources, and validation that catches problems before consumers do.
• Support experimentation and measurement design as we roll out platform and product changes.
• Grow into quantitative optimization work, including improving how we model and select what offers and experiences we serve to each user session.

• 3 to 5 years of experience in analytics, analytics engineering, or data science.
• Hands-on experience building with LLMs against real data, including how you validated whether the output was trustworthy. Demos and prompt experimentation are not sufficient.
• Production experience with dbt: you have owned models, written tests, worked in Git, and shipped.
• Advanced SQL and working proficiency in Python.
• Experience with a cloud data warehouse (BigQuery preferred, Snowflake or Redshift acceptable).
• Practical experience with data quality work: reconciliation across sources, and validating your own output.
• The ability to hold a business conversation. You should be able to walk us through a question you chased end to end, what you thought was happening, what you found, and what changed as a result.
• Comfort with ambiguity and a track record of deciding what to work on rather than waiting to be told.

• Experience in AdTech, performance marketing, or marketplace businesses.
• Experience as the first or only data person at a company.
• Product analytics experience: success metrics, event instrumentation, and measuring feature impact.
• Experimentation design and applied statistics.
• Exposure to bid optimization, yield management, or similar decision-modeling work.
• Familiarity with monitoring and observability tooling such as Datadog or Grafana.

• Self-Starter – You take initiative and solve problems before they become problems.
• Curious & Analytical – You ask why, dig deep into data, and experiment to learn what works.
• Efficiency-Oriented – Fast, focused, and action-driven.
• Relationship-Focused – You build trust and deliver value.
• Organized Multitasker – You thrive in fast-paced, high-velocity environments.
• Strategic Communicator – You translate performance into insight and insight into action.