Google Cloud Platform Data Engineer

Role — Google Cloud Platform Data Engineer

Location: Remote

Duration: 12+ months

Engagement: SAS to Google Cloud migration for a US healthcare client Duration: 12 months

Role summary:

You will rebuild an established on-premise analytics estate on BigQuery — converting existing batch processing logic into SQL, building the ingestion and transformation pipelines, and proving the new outputs match the old ones exactly.

The work is layered: raw data lands as it arrived, is conformed and deduplicated, then modelled for reporting. Transformations are written in SQL and held in Git, reviewed by pull request. Non-SQL work — file format conversion, XML parsing, file generation — runs as containerised Python.

Must have
10-15 years in data engineering overall.
4+ years hands-on with Google Cloud, with BigQuery as the primary platform.
Advanced SQL. This is the core of the role. Window functions, complex multi-table joins, set-based thinking, and the discipline to write SQL that is readable and reviewable by someone else six months later.
BigQuery in depth. Partitioning and clustering strategy, query cost control, slot behaviour, incremental and merge patterns, nested and repeated fields, and the practical difference between a query that costs cents and one that costs hundreds of dollars.
Dataform or dbt. Dependency-driven SQL transformation frameworks, incremental models, assertions and testing, and development through Git with pull request review.
Python to a production standard, and containerising it for Cloud Run.
Dimensional and layered data modelling — raw, conformed and consumption layers, grain and key design, slowly changing dimensions, and handling history correctly.
Reconciliation and data quality. Proving a migrated output matches the source system — row counts, checksums, distinct key counts, boundary values — and investigating when it does not.
Git and pull request workflow as a normal way of working.
Client-facing communication. These roles sit with the client team daily and are expected to explain design decisions, run working sessions and represent the solution in architecture and security reviews. Written and spoken clarity matters as much as technical depth.
Seniority to own a design decision and defend it. At this level you are expected to set and enforce the SQL and modelling conventions the team works to, review others' work, and hold a position in front of the client's data modeller and data product owner.

Strongly preferred:
Ability to read SAS — Base SAS, DATA steps, PROC SQL, macros. You do not need to be a SAS developer, but being able to read a program and understand what it produces is a significant advantage on this engagement and will be weighted heavily.
Experience migrating from a legacy analytics platform to a cloud warehouse.
Handling PHI, PII or equivalently regulated data, including masking and de-identification.
Looker and LookML.
File format work — Parquet, Avro, fixed-width and delimited files, XML parsing at volume.
Cloud Workflows, Cloud Scheduler, Pub/Sub.
Terraform literacy, enough to make and raise routine infrastructure changes for review.

What you will deliver
Ingestion pipelines for file-based and database sources.
Conversion of existing batch programs into SQL models, preserving business logic exactly.
Reconciliation evidence for each migrated output, demonstrating parity with the current system.
Operational pipelines with logging, alerting and data quality checks, documented well enough to be supported by someone who did not build them.
Screening notes for the recruiter

SQL depth is the strongest predictor of success in this role — more than the number of cloud services a candidate can name. Ask for a description of the most complex transformation they have written and why it was difficult. A candidate who has only moved data between systems without owning the business logic is not a fit.

SAS reading ability should be treated as a tiebreaker in the candidate's favour wherever it appears, even if the experience is several years old. Candidates at this experience level frequently worked on SAS, Teradata or Informatica earlier in their careers — that background is an asset on this engagement, not a gap.

Note on the experience range: the 10-15 years is total data engineering. Four or more years hands-on with Google Cloud Platform is the bar; earlier years on another platform are expected and welcome.

Thanks & Regards