Senior Analytics Engineer — Data Delivery

Senior Analytics Engineer, Data Delivery

XtendOps · Business Intelligence / Data · Full-time · Remote

ABOUT XTENDOPS

XtendOps delivers outsourced customer experience and back-office operations for a growing portfolio of client accounts. Our Business Intelligence / Data team owns the warehouse, the semantic layer and the delivery path behind the reporting and in-product analytics that executives, team leaders and clients use to track performance, monitor financials and find root causes.

ABOUT THE ROLE

Our platform is already in production: a Snowflake warehouse with a medallion-style architecture, transformation in dbt, orchestration in Airflow, and a metric catalog held as data rather than code. That modeled output is becoming the data layer behind a client-facing analytics product. The work ahead is bringing our portfolio of accounts onto it.

This role is weighted toward the last mile. Not just modeling a metric correctly, but conforming it to shared definitions, proving it against a known-good source, wrapping it in a contract, and getting it into the product unchanged and on schedule.

This is a hands-on senior IC role. You will own production models and the feeds built on them, and set the standard others model to.

Reports to: Senior Data Engineer

Works with: Data engineers, analysts, BI leads, product and platform engineers, PM, operations and client stakeholders

Location: Remote, with overlap with North America and Asia-Pacific business hours

WHAT YOU'LL DO

Product-ready data delivery

• Take accounts from modeled warehouse output to live in the analytics product: conform metrics, build the serving feed, reconcile it and schedule delivery

• Own the metric conformance layer that maps client-specific source values to master, catalog-registered definitions

• Design and maintain feed contracts: grain, keys, period semantics, null and late-arrival behavior, versioning and backfill rules

• Build automated reconciliation that runs on every load and proves numbers arrive in the product unchanged

• Build serving models that support account-, team- and agent-level entitlements and filtering

• Own freshness and delivery SLAs for your feeds, including monitoring and alerting

• Instrument the delivery path so a wrong tile can be traced to a run, model and source in minutes

Client onboarding at scale

• Move accounts from source data landing through conformance, reconciliation and a scheduled, monitored push

• Assess new and legacy accounts honestly and defend realistic effort estimates

• Migrate legacy reporting assets onto the shared foundation instead of carrying them forward

• Improve onboarding scaffolding so every new account is measurably cheaper than the last

Shared metric foundation

• Build and maintain core metric models across accounts: handle time, CSAT, QA, FCR, SLA attainment and workforce measures

• Own master and conformed dimensions (company, client, employee, team, channel, date)

• Maintain the master metric dictionary with the analyst team

• Resolve definition conflicts between accounts and document deliberate divergence

• Collapse per-account metric copies into shared models

Quality and correctness

• Write tests covering structure, business logic and cross-model consistency

• Apply model contracts and versioning so breaking changes are caught in CI, not in a client dashboard

• Own freshness and anomaly monitoring and lead incident response and post-mortems

• Treat incentive, compensation and billing reporting as audit-grade work

Collaboration and delivery

• Review analyst modeling contributions and raise the quality bar

• Turn spreadsheet-and-conversation requirements into buildable, testable specs

• Partner with product and platform engineers, and push back when a request would break a shared definition

• Mentor mid-level analytics engineers and analysts

• Write documentation, runbooks and architectural decision records

• Work in Agile sprints with a definition of done that includes tests, reconciliation and docs

WHAT YOU'LL NEED

Required

• 5 to 8 years in analytics engineering, data engineering or BI engineering with substantial production data modeling

• Hands-on production dbt experience: models, sources, tests, macros, exposures, contracts, versions, materializations and incremental strategies

• Strong SQL, including window functions, complex aggregation and query performance reasoning

• Proven delivery of data to a downstream consumer under contract (dataset, API feed, reverse-ETL or serving table)

• Experience building automated reconciliation checks and explaining discrepancies

• Solid dimensional modeling: facts, dimensions, grain, conformed dimensions, surrogate keys

• Git and pull-request workflow, comfortable giving and receiving SQL review

• Enough Python to read, modify and extend ingestion and utility code

• Ability to read a pipeline dependency graph and understand downstream impact

• Comfort with ambiguity and good judgment on when to ask versus propose

• Clear written English for definitions, documentation and specs

Strongly preferred

• Snowflake, including warehouse cost awareness

• Multi-client or multi-tenant modeling

• Metadata- or catalog-driven modeling

• Experience feeding a customer-facing analytics product or embedded dashboards

• Row- and column-level security or multi-tenant access control

• Apache Airflow (Astronomer or equivalent): dependencies, retries, SLAs, backfills

• Contact center, BPO or WFM metrics: AHT, CSAT, QA, FCR, SLA, adherence, occupancy

• Slowly changing dimensions, especially for employee and roster data

• Ingesting from imperfect sources like spreadsheets and manual trackers

• Financially consequential reporting with audit trails

• Migrating legacy BI assets (Looker, Tableau) onto a governed modeling layer

• Medallion-style warehouse architecture

• Using AI tools like Claude in daily engineering work, with the judgment to verify output

Not required: cloud provisioning, CI infrastructure or front-end code. You do need to understand where your work sits in the wider pipeline.

HOW SUCCESS IS MEASURED

• Accounts you own go live in the product within the estimated window and stay live

• Every live feed has a green automated reconciliation

• Breaking changes are caught in CI and downstream consumers are never surprised

• Core metrics come from shared models, not per-account copies

• Freshness targets are met and incidents resolved within agreed response times

• Onboarding cost per account trends down over time

• Estimates hold up against actual delivery

• Analysts are unblocked, not queued behind you

• Documentation is current enough that someone else can operate what you own

OUR STACK

Snowflake · dbt · Apache Airflow on Astronomer · AWS (GCP secondary) · GitHub monorepo with CI testing and SQL linting · ClickUp · Tableau · Claude

WHY THIS ROLE MATTERS

Modeling a metric is no longer the hard part. The hard part is everything between a correct model and a number an executive trusts on screen: one shared definition, a contract the product can build on, reconciliation that proves nothing changed in transit, and a schedule that holds.

Few accounts have made that journey end to end so far. This role is how the rest get there, and how onboarding a client becomes configuration rather than a project. In several accounts these numbers feed agent incentive calculations, where correctness is not negotiable.

The foundation is being designed now, not maintained.

Compensation, benefits and employment details will be shared during the interview process.