DEPARTMENT: BUSINESS INTELLIGENCE / DATA
ABOUT XTENDOPS
XO delivers outsourced customer experience and business operations, built on operational excellence, AI-driven innovation, and a bold approach to problem-solving. We help companies run smarter, operate more efficiently, and drive measurable business impact.
We are in the middle of a massive evolution, expanding beyond traditional outsourcing into a next-generation model that integrates AI, automation, and SaaS-driven solutions to deliver outcome-focused business services. Our vision is not just to improve CX. It is to reinvent how businesses leverage technology, data, and human expertise to create intelligent, scalable, and revenue-generating service models.
MAIN JOB OBJECTIVE: The Data Platform Engineer owns the infrastructure and orchestration that the data platform runs on, and builds the deployment path that lets new client accounts be stood up without manual provisioning. Orchestration and CI are already in place, but infrastructure is not yet managed as code and platform ownership currently sits with the team lead alongside other responsibilities. This role is central to scaling client onboarding into a repeatable process and to building the delivery path that pushes modeled data into a client-facing analytics product. This is a hands-on engineering role with broad ownership of the platform layer, not of production data models or application front ends.
RESPONSIBILITIES AND MAIN ACTIVITIES
Orchestration:
- Own Apache Airflow and the orchestration layer, including the factory pattern that generates pipelines per client, and extend it as accounts are added.
- Build and maintain custom operators and shared pipeline components, so common patterns are solved once rather than per account.
- Own retry behavior, idempotency, and backfill strategy, so a failed run recovers cleanly instead of leaving partial data behind.
- Sequence and schedule the full path from ingestion through transformation to publishing, coordinating with the engineers who own the models.
Infrastructure as Code and Provisioning:
- Build the infrastructure-as-code foundation, which is the prerequisite for repeatable client onboarding and does not yet exist.
- Own environment management across development, staging, and production, including for the serving path.
- Build the per-client provisioning path so that standing up a new account is a configuration change and a pipeline run rather than a manual runbook.
- Own secrets and credential management as a managed practice.
Ingestion Mechanics:
- Own the mechanics of scheduled extraction from spreadsheets, shared drives, and client APIs, with reliability handled once in shared components rather than reimplemented per pipeline. –
- Build ingestion that fails loudly and recovers predictably, since a meaningful share of sources are manual trackers that change without notice.
- Work with the analytics engineers on the contract between raw landed data and the modeling layer.
Serving Path:
- Own the infrastructure behind the delivery path that pushes modeled data into the internal software platform, including its monitoring and its data freshness guarantees.
- Implement and enforce tenant isolation in the serving layer, so per-client data separation is a property of the platform rather than of each query written against it.
- Partner with the product and application engineers who consume the data on the contract between platform output and their application.
Observability, CI, and Cost:
- Own CI and CD for the data platform, including how pipeline and infrastructure changes are gated safely. – Build observability that answers whether data actually landed, not only whether a job exited zero, with alerting that distinguishes a real failure from noise.
- Own cloud spend visibility across the data stack: budgets, usage alerts, quotas, and attribution of cost by account and initiative.
- Take part in incident response and post-mortems, and turn findings into backlog items.
Documentation and Sprint Delivery:
- Write and maintain runbooks and platform documentation to the standard that someone else can operate and extend what is built.
- Work within the team's sprint cadence, and contribute to architectural decision records when a platform decision is worth writing down.
QUALIFICATIONS AND EXPERIENCE
- 3–5 years in data engineering, platform engineering, DevOps, or a comparable infrastructure-focused role.
- Airflow in production required: custom operators, dynamically generated DAGs, deliberate handling of retries and idempotency, and backfills planned to be right the first time.
- Python to a maintainable standard, writing shared code that others depend on and extend.
- Infrastructure as code required. Terraform preferred, another tool acceptable if the fundamentals are solid. This capability does not exist on the team today.
- CI and CD design, including how to gate data pipeline and infrastructure changes safely.
- Secrets and credential management as a discipline.
- Observability instincts: instrumentation that surfaces silent failure, and alerting that people do not learn to ignore.
- Debugging across system boundaries, since most failures live between a client source, the warehouse, and the serving path rather than inside any one of them.
- Working data literacy: ability to read a pipeline dependency graph, understand what a data model is and what depends on it, and write enough SQL to diagnose a pipeline problem.
- Clear written communication in English.
PREFERRED
- Snowflake administration: warehouses, roles, resource monitors, and cost attribution.
- Cloud cost management or FinOps exposure: budgets, spend attribution, usage monitoring, and treating refresh frequency as a cost decision.
- Operational database experience, particularly tuning a read-heavy store used to serve an application.
- Semantic layer or metrics API experience, such as Cube or a comparable tool.
- Multi-tenant infrastructure and per-tenant isolation, including row-level or schema-level separation and the trade-offs between them.
- Ingestion from unreliable sources such as client spreadsheets and manual trackers.
- Containers and container-based deployment.
- dbt, at the level of running and orchestrating it and being able to read models.
- Enough understanding of web application delivery (APIs, caching, deployment) to coordinate a data-to-application delivery path.
- AWS, and secondarily Google Cloud.
- Experience introducing infrastructure as code or observability to a team that had neither.
AVAILABILITY
- Full-time, remote, with overlap required across North America and Asia-Pacific business hours.
SKILLS
- Apache Airflow and Pipeline Orchestration
- Infrastructure as Code (Terraform or Equivalent)
- Python (Shared, Maintainable Code)
- CI/CD Design and Gating
- Secrets and Credential Management
- Observability and Alerting Design
- Cross-System Debugging
- Cloud Cost Management and Attribution
- Multi-Tenant Architecture and Isolation
- Documentation and Runbook Writing
- Agile/Scrum Sprint Delivery