AI Analytics Engineer

REP Fitness•Westminster, CO

18h•$120,000 - $140,000•Hybrid

About The Position

REP is hiring a senior individual contributor to build trusted analytics data products and the semantic and context layers that allow employees and AI agents to query enterprise data accurately. The role works with REP’s Microsoft Fabric data platform, internal conversational BI platform ChatDash, Claude, REP MCP, and other sources. It owns metric definitions of record and develops pipelines using a medallion architecture, while serving as the technical owner of an external data engineering partner and collaborating with the AI Solutions Engineer, who builds agents and tools. A central part of the role is hardening analytics that business users have already built in Claude or ChatDash. The engineer reverse-engineers the questions, logic, sources, and implied metric definitions in those working versions, then moves their data flows from ad hoc extracts and queries to governed pipelines and certified semantics. The hardened solution should preserve the business user’s working version, answer the same questions, and make trusted data and definitions accessible wherever people build. The role builds and tests Fabric pipelines and gold data products across bronze, silver, and gold layers; establishes a certified semantic layer; and creates evaluation frameworks and guardrails for AI-generated answers and agent data access. It uses AI coding assistants to accelerate development while applying human judgment to specifications, reviews, testing, accuracy, and governance. The position is based at REP’s headquarters in a hybrid environment, with at least three office days per week and flexibility in which days and hours.

Requirements

Bachelor’s degree from a four-year college or university.
At least 5 years of experience in analytics engineering, data engineering, or a closely related role, operating as a senior individual contributor.
Ability to write high-performance SQL and Python dataflows against a cloud warehouse.
Experience building and maintaining models in a medallion (bronze/silver/gold) architecture with dbt-style transformation, testing, and documentation discipline.
Experience designing or owning a semantic or metrics layer as the single source of truth for business metrics.
Experience using AI coding assistants to materially accelerate data pipeline delivery.
Experience working across complex source systems without having the data model supplied.
Experience implementing row- and column-level security, access policy, and lineage in a production data platform.
Experience turning analyst- or business-built prototypes into production data products.
Experience directing external data engineering partners through specifications and acceptance testing.
Ability to interpret and harden business users’ working prototypes without taking ownership away from the builders.
Ability to distinguish problems belonging in the semantic layer from those belonging in prompts, application logic, or ad hoc queries, and to identify definitional, grain, and source-system problems.
Verification-minded approach: test AI-generated output before trusting it and communicate plainly when confidence is low.
Ability to use AI coding assistants as tools under review rather than as authorities.
Ability to apply data governance independently and make conservative calls when data exposure is ambiguous.
Ability to translate business needs into executable vendor specifications and reject work that does not meet requirements.

Nice To Haves

Microsoft Fabric experience is preferred; Snowflake, Databricks, or BigQuery experience is transferable.
Experience building or contributing to evaluation frameworks for LLM or text-to-SQL accuracy, or exposing governed data to AI agents through MCP or similar interfaces, is a strong plus.

Responsibilities

Reverse-engineer business-built Claude and ChatDash solutions to understand their questions, logic, data sources, and implied metric definitions.
Move data behind business-built solutions into governed Fabric pipelines and certified semantics, replacing ad hoc extracts, uploads, and one-off queries while preserving the solutions’ usefulness.
Make certified data and semantics accessible through Claude, ChatDash, REP MCP, and other development approaches.
Work directly with business builders to confirm hardened solutions answer the same questions as their working versions and explain any differences.
Build end-to-end Fabric pipelines and gold data products, including ingestion into bronze, cleansing and conforming in silver, and gold tables and views for reporting and agents.
Generate test suites and automated reconciliation against source-system reports to reduce manual user acceptance testing.
Derive specifications and acceptance criteria from existing business-built solutions, and confirm freshness targets and a business owner before productionizing.
Keep transformation code version-controlled, tested, and documented so other engineers or agents can extend it.
Deliver a certified semantic layer over the Fabric gold layer, including metrics, entity definitions, join paths, and grain rules, beginning with inventory and net sales.
Establish one definition of record per metric and retire or explicitly map conflicting definitions across Power BI, Polar, and NetSuite.
Structure business rules, metadata, and documentation as durable working memory for agents.
Resolve data-trust escalations at their root in definitions and lineage rather than through one-off reconciliations.
Create an evaluation framework with a question bank, known-correct answers, and automated regression checks for semantic-layer and model changes.
Publish regular accuracy scorecards for leaders and the AI Governance committee, identifying where agent answers can and cannot be trusted.
Catch answer-quality regressions before release.
Implement access controls and row- and column-level policies in semantic and MCP layers so agents return only data the requesting user is authorized to see.
Define which agent actions are read-only or write-capable and review write paths against REP’s connector governance framework before exposure.
Maintain auditable lineage from agent-delivered answers to source data for Finance and governance review.
Serve as technical owner of the external data engineering partner by writing measurable build specifications, reviewing deliverables against acceptance criteria, and rejecting work that does not meet the bar.
Keep vendor scopes bounded to vertical, testable work so REP retains understanding of its data.
Use AI coding assistants to accelerate pipeline development while reviewing and testing generated work before it ships.
Apply governance to the role’s own work, make conservative decisions about ambiguous data exposure, and uphold architecture gates.

Benefits

Medical, dental, and vision insurance.
Health Savings Account with employer contributions.
Flexible Spending Account options.
Long- and short-term disability, life, and AD&D insurance, with premiums fully covered by REP.
401(k) and Roth options with employer match up to 4%.
Employee Assistance Program.
Four weeks of PTO accrued during the first year.
48 hours of sick leave annually.
Seven paid company holidays and one additional floating holiday.
Paid parental leave beginning on day one.
Paid birthday off.
Bonus structure tied to individual and company-wide performance.
Full access to the employee gym.
Employee discount on REP equipment, apparel, and nutrition.
Daily snacks and drinks for onsite employees.
Company and team-building events.
Professional growth and development within the organization.
Flexible hybrid work environment and flexibility in workdays and hours.