Analytics Engineer, Integrity

National Basketball Association (NBA)•New York, NY

12d•$130,000 - $150,000•Hybrid

About The Position

The NBA’s Basketball Strategy & Growth department is seeking an analytics engineer to own the ingestion and transformation of the data that powers the Integrity Team’s work on the league’s gaming policy. The role is responsible for turning disparate and inconsistent external feeds — data from betting partners and operators, open-source intelligence (OSINT), and third-party vendors — into well-modeled, tested, and documented datasets, and for integrating those feeds with internal league basketball data so that betting activity can be evaluated in the context of what happened on the court. These models are the foundation for downstream data science and machine learning work: the anomaly detection, alerting, and AI-driven software the Integrity Team uses to surface conduct that may violate the gaming policy. The analytics engineer joins a team of data scientists, engineers and basketball experts and collaborates closely with the Legal department on any investigation relating to a potential violation of the gaming policy. The work is confidential, frequently time-sensitive, and expected to hold up to legal scrutiny — lineage, definitions, and data quality controls matter as much as speed.

Requirements

Advanced SQL skills with demonstrated experience developing complex analytical data models.
Production experience with analytics engineering frameworks such as dbt, including testing, documentation, macros, and managing a large model DAG.
Production experience with Snowflake, including performance and cost management such as warehouse sizing, clustering, and query tuning.
Strong understanding of dimensional modeling, data warehousing, and analytics engineering best practices.
Understanding of sports betting markets — odds, line movement, limits, player props, and market maker behavior — and of basketball data such as play-by-play, box score, and tracking data.
Proficiency in Python for ingestion, API integration, and data processing.
Experience integrating data from multiple external partners or vendors, including messy, high-volume, semi-structured, and unstructured sources.
Experience building and operating orchestrated data pipelines, including scheduling, dependency management, retries, and monitoring.
Experience developing semantic layers, standardized business definitions, and reusable analytical datasets.
Experience implementing automated data quality testing, monitoring, and documentation to support trusted analytical data assets.
Experience building pipelines that serve downstream data science and machine learning consumers.
Familiarity with Git-based development workflows (GitHub or Azure DevOps), code review, and CI practices.
Strong communication and stakeholder management skills, with the ability to translate business and investigative requirements into scalable analytical solutions for both technical and non-technical audiences, including attorneys and investigators.
Sound judgment and discretion in handling confidential and legally sensitive information.
Confident individual, able to work in a fast-paced environment and manage short-term deliverables along with long-term projects.
Excels when working as part of a team, collaborating effectively both within BSG Integrity and cross-functionally.

Nice To Haves

Experience with Databricks, including Spark, Delta Lake, Unity Catalog, and/or ML lifecycle tooling such as MLflow.
Experience with Airflow, Dagster, Prefect, or a comparable orchestration platform.
Experience with entity resolution, identity matching, or record linkage.
Familiarity with MLOps practices such as feature pipelines, model registries, training/retraining workflows, and model monitoring.
Prior work on integrity, surveillance, fraud, AML, or trust and safety data is a plus.
Experience in sports, gaming, or another regulated or compliance-driven environment is a plus.

Responsibilities

Own the ingestion of data from disparate betting partners, operators, and data vendors — each with its own schema, granularity, delivery cadence, and idiosyncrasies — and normalize them into a consistent, conformed model of markets, wagers, prices, and accounts.
Build and maintain ingestion pipelines for OSINT and other unstructured or semi-structured sources, including the entity resolution work required to link external identities to known accounts and subjects.
Integrate betting and OSINT data with internal league basketball data — schedule, play-by-play, box score, tracking, officiating, and player availability data — so that activity can be analyzed in game context.
Develop production-ready analytical data models in dbt on Snowflake, applying dimensional modeling and analytics engineering best practices, including layered staging and mart designs, incremental models, and consistent naming conventions.
Create curated, reusable pipelines and feature-ready datasets that support downstream data science and machine learning work, including model development, backtesting, and production inference.
Implement automated data quality testing, freshness and volume monitoring, and reconciliation checks on incoming feeds so that vendor changes, outages, or silent data loss are detected before they affect investigative output.
Build and manage orchestration and scheduling for pipelines, including dependency management, retries, alerting, and service levels appropriate to time-sensitive feeds.
Develop and maintain a semantic layer with standardized business definitions and metrics so that analysts, data scientists, and the Legal department work from a single version of the truth.
Document models, sources, lineage, and assumptions to support knowledge transfer, auditability, and the evidentiary needs of investigations.
Onboard new betting partners and data providers — evaluating feed quality, defining requirements and specifications with providers, and absorbing vendor schema changes without breaking downstream consumers.
Partner with data scientists to translate analytical and modeling requirements into scalable data models, and with the full stack engineer to expose those datasets to the platform’s backend jobs and UI.
Apply appropriate access controls, data classification, and retention practices to sensitive betting, personal, and investigative data.