Staff Analytics Engineer

Job Description

Flex is transforming the traditional, fragmented $90B+ moving, storage, and last-mile delivery industry into a seamless, technology-driven experience. As Staff Analytics Engineer, you will own how this company measures itself: the data models, the metric definitions, and the reporting that Finance, Operations, Sales, Marketing, and other functions use to make decisions every day.

This is a full-stack data role at the senior individual-contributor level. You will sit directly with stakeholders to work out what decision they are trying to make, build the models and reporting that answer it, and defend the numbers when they land in front of executives. You will be the senior-most analytics practitioner outside of leadership, with the autonomy that implies and the accountability that comes with it. This role is open to USA remote, with need-based travel to our offices. Regardless of location, you'll be expected to work ET hours.

How the Work Splits:

\n
• ~35% data modeling : designing and building dbt models on BigQuery.
\n
• ~35% stakeholder analysis and reporting : requirements, metric definitions, and the Sigma dashboards and analyses that deliver them.
\n
• ~20% ad-hoc investigation : the questions that don't have a dashboard yet, and the data-quality problems that surface underneath them.
\n
• ~10% ingestion and platform : Fivetran connectors, lightweight Python pipelines, testing, documentation, and production reliability.
\n

Core Responsibilities

\n
• Stakeholder Ownership: Be the direct point of contact for Finance, Operations, Sales, Marketing, and other functions. Turn vague requests into specifications, surface the questions only they can answer, and drive them to a decision.
\n
• Data Modeling: Own the design and layering of our dbt project on BigQuery (staging through marts) including tests, documentation, and the production runbook.
\n
• Metric Definitions: Decide and document how Flex measures revenue, contribution margin, utilization, CAC, and the rest. These definitions go into board reporting and pricing decisions, so the reasoning behind each one is part of the deliverable.
\n
• Reporting: Build the analytics experiences our teams actually use in Sigma, and own the platform behind them, the data dictionary, governance, permissions, and the discipline to retire what nobody should be using anymore.
\n
• Data Ingestion: Manage our Fivetran connectors and build the pipelines Fivetran can't.
\n

Key Requirements

\n
• Depth in Data: 10+ years working in data: analytics engineering, analytics, or data science, including real time as the senior-most data person where the definitions and the correctness were yours to defend.
\n
• Excellent SQL: Comfortable with window functions, semi-structured data, incremental logic, and debugging a query whose result you don't believe.
\n
• Production dbt Experience: You have designed a project's layering, not just added models to someone else's.
\n
• BI Ownership: You have owned a modern BI platform end to end — Sigma, Looker, Tableau, Power BI, Omni, or Hex — including the governance, not just the charts. Sigma specifically is a plus, not a requirement.
\n
• Stakeholder Fluency: You can run the meeting, push back on a badly formed request, explain a methodology choice to a CFO in their language, and get a decision out of people who would rather not make one.
\n
• Logistics or Marketplace Domain Experience: You have done data work inside a business that moves physical goods or people — logistics, moving, delivery, field services, fleet, or a two-sided labor marketplace. You understand how operational reality shows up in the data: dispatch and routing, crews and capacity, no-shows and reschedules, jobs that get split or cancelled halfway through.
\n
• Comfort in Messy Operational Data: Our sources include a document database, payments, a CRM, payroll, ad platform data, spreadsheets, and a host of other sources. Data quality varies, and the gap between what the data says and what happened on the ground is frequently the actual finding.
\n
• Root-Cause Discipline: When something looks wrong, you trace it to a proven cause in the code or the data rather than shipping the most plausible explanation.
\n
• Python: Sufficient for ingestion scripts and analysis: pandas, an API client, a notebook.
\n
• Education: A Bachelor's degree in a quantitative, engineering, or business field; advanced degrees are a bonus, not a requirement.
\n

Nice to Have

\n
• Financial and unit-economics literacy : revenue recognition, cash versus accrual, contribution margin, CAC.
\n
• Applied ML or statistics: forecasting, pricing, propensity, causal inference. As the analytics function matures under you, there is room to contribute to data science and machine learning projects.
\n
• Experience standing up a data platform's unglamorous essentials: CI, environment separation, backups, cost control.
\n