Sr. Data Engineer

TKO Group Holdings, Inc. is a premium sports and entertainment company whose portfolio includes UFC, WWE, PBR, IMG, and On Location. The Senior Data Engineer, Data Platform will own and improve shared data pipelines, models, Snowflake infrastructure, ingestion, orchestration, governance, and reliability while helping unify TKO’s data platforms. The role also supports analytics, machine learning, and AI workloads and mentors engineers and delivery partners on platform standards.

Responsibilities:

• Own pipelines, data models, and platform reliability end to end, covering all ingress and egress paths into and out of Snowflake
• Act as an escalation point for platform incidents — triage orchestration, transformation, and monitoring alerts, run root cause analysis, and close issues out with permanent fixes rather than recurring manual intervention
• Define and hold service levels for the pipelines the business depends on: freshness, job success rate, and time to recover. Report honestly against them
• Provide debug, triage, support, and design guidance to engineers and analysts working anywhere in the platform
• Share in the team's on-call and event-support rotation, including heightened coverage around major live events
• Design and build the shared components of TKO's unified platform: the canonical medallion layering, the role and access taxonomy, the shared ingestion library, and the orchestration framework every business unit will run on
• Lead Snowflake consolidation work as the separate accounts converge — cross-region replication, account merges, re-pointing secure shares, parity validation, and cutover
• Retire duplicated tooling across the estate rather than relocating it: self-managed Airflow, notebook-driven ingestion, SSIS, and legacy CI/CD
• Build golden paths — templates, reference implementations, and documentation — so business unit teams and delivery partners can onboard a source or ship a model without a ticket to this team for every change
• Extend and maintain a shared Python ingestion library: reusable connectors for SFTP, database, API and file sources; S3 landing with Snowflake COPY and Snowpipe sinks; watermarking and incremental logic; registry and observability hooks
• Run managed ingestion (Fivetran) as the default for supported SaaS sources, and make the build-versus-buy call where no connector exists
• Own Prefect on EKS — deployments, work pools, worker health, containerized flows — and the standard flow taxonomy the whole organization writes against
• Consolidate connector logic that exists today into a shared collect/ingest library, instead of maintaining three versions of the same capability
• Administer Snowflake as a shared, multi-tenant platform: warehouse sizing and workload isolation, resource monitors, query and clustering optimization, replication, and secure sharing with external partners
• Manage Snowflake objects declaratively in Terraform so that structure and access are reviewable, reproducible, and auditable rather than accumulated through ad hoc SQL
• Own warehouse cost. Find the queries, models, and schedules driving consumption, fix them, and give leadership a defensible view of where credits are going
• Build and maintain AWS infrastructure with Terraform across the TKO accounts — EKS, networking, IAM, S3, Lambda, and secrets management — operating at the workload layer inside central IT's guardrails rather than standing up a parallel estate
• Maintain CI/CD on GitHub Actions with AWS OIDC and no static credentials, including the dev, UAT, and production promotion path, required-reviewer gates, and the dbt tests and data quality checks that block a bad change before it lands
• Raise the engineering bar through code review, testing standards, and documentation that holds up after the author moves on
• Build and review dbt models against the canonical medallion, and merge the organization's separate macro libraries into one shared set — identity resolution and feature engineering from the consumer side, policy generation from the brand side
• Partner with Analytics Engineering and Data Science so semantic models, feature pipelines, and reporting sit on governed, versioned datasets rather than one-off extracts
• Implement masking, tokenization, and row-access policies as code across the estate — encrypted at landing, decrypted at the marts — so PII is handled the same way everywhere it lands
• Maintain network policies, secrets and key rotation, and SSO-driven role provisioning; support access reviews, audit requests, and remediation of security findings
• Support cataloging and lineage tooling, and build to the standards set by TKO's Master Data Management and Governance function
• Keep the platform ready for the workloads built on top of it: in-platform model training and serving, feature pipelines, and the curated, retrieval-ready datasets that AI applications depend on
• Use AI-assisted engineering tooling in daily development and help establish sensible team practice around it
• Work directly with the global delivery partners — mentoring them onto the target stack through their migration waves and reviewing their work against platform standards
• Mentor engineers on the team and serve as a technical reference point on design decisions, without formal management responsibility
• Explain technical trade-offs plainly to stakeholders who are not engineers, and to leadership

Requirements:

• 7+ years building and operating production data platforms, including at least 3 years weighted toward shared platform and infrastructure work rather than single-team pipeline delivery
• Deep hands-on Snowflake experience at an administrator level: RBAC design, warehouse sizing and cost management, masking and row-access policies, replication, secure data sharing, and performance tuning
• Strong Python and SQL, with the ability to write and maintain a shared library other engineers build against — not only scripts and DAGs
• Production experience with a modern orchestrator (Prefect, Airflow, or Dagster), including running and upgrading it on Kubernetes or comparable infrastructure rather than only consuming a managed service
• Hands-on infrastructure as code with Terraform on AWS: EKS or ECS, IAM, VPC networking, S3, Lambda, and secrets management
• dbt at scale — project structure, macros, testing, and the judgment to consolidate overlapping projects rather than accumulate them
• Ownership of CI/CD for data systems, GitHub Actions preferred, including keyless OIDC authentication, multi-environment promotion, and automated quality gates
• Experience with managed ingestion (Fivetran or comparable), CDC, Snowpipe, and event or behavioral data platforms such as Amplitude or Segment
• Operational discipline: observability tooling such as Datadog, alerting design, incident response, and post-incident follow-through
• Demonstrated work on a platform migration or consolidation of real scope — merging warehouse accounts, replacing an orchestrator, or retiring a legacy stack — including the parity validation and cutover planning that determines whether one succeeds
• A record of setting standards other engineers actually adopt: reusable frameworks, reference implementations, documentation, and code review
• Comfortable working alongside offshore and partner delivery teams and raising their output to the platform's standards
• Clear written communication, particularly design documents, runbooks, and incident write-ups
• Practical experience meeting data governance requirements — PII classification, access controls, retention, and lineage — preferred
• Exposure to ML or AI workloads running on the data platform, such as in-warehouse training and serving, feature stores, or retrieval pipelines, preferred
• Experience in media, sports, entertainment, live events, ticketing, or consumer and fan data environments preferred