Key Responsibilities :

- Independently write complex SQL queries in Snowflake to validate analysis findings and reconcile outputs with source data.

- Perform end-to-end data-quality validation across source, transformation, and target layers.

- Develop and execute dbt schema tests and data tests, including uniqueness, not-null, referential-integrity and accepted-value validations.

- Identify duplicates, invalid or expired records, inconsistencies, missing data and false positives in detection logic.

- Perform detailed reconciliation using row counts, aggregates, source-to-target comparisons and edge-case validations.

- Validate data pipelines, transformation logic, analytical datasets and downstream outputs.

- Investigate data-quality issues, identify root causes and work with data engineering/analytics teams on resolution.

- Maintain clear evidence and documentation of test scenarios, findings, reconciliation results and defects.

- Where applicable, validate AI/ML or LLM-generated outputs by sampling against ground truth, measuring accuracy, identifying failure patterns and checking consistency across repeated runs.

- Support data lineage, traceability and data-cataloguing requirements.

Required Skills :

- 3-7 years of relevant experience in data testing, data validation, data engineering QA or analytics quality assurance.

- Strong hands-on SQL skills with the ability to independently investigate and reconcile data.

- Practical experience working with Snowflake.

- Hands-on experience with dbt testing, including schema tests, data tests and assertions.

- Strong understanding of data-quality dimensions and validation techniques.

- Experience identifying duplicates, invalid/expired records, inconsistencies and false positives.

- Strong reconciliation skills, including source-to-target validation, row counts, aggregates and edge cases.

- Experience testing data pipelines and analytics outputs, rather than primarily UI/application testing.

Preferred Skills :

Experience validating AI/ML or LLM-generated results is strongly preferred, particularly ground-truth comparison, accuracy measurement, failure-pattern analysis and consistency testing. Familiarity with Snowflake tooling and exposure to data cataloguing, lineage and traceability will also be valuable.

Knowledge of supply chain, ERP/SAP or regulated data environments is an advantage, especially the ability to identify outputs that appear technically plausible but are incorrect from a business perspective. Familiarity with Dagster or similar orchestration tools is also desirable.