- Deep Python expertise including PySpark, pyarrow, and an understanding of Spark's execution model (driver vs executor, serialization constraints, partition fan-out)

- Hands-on experience with at least one SaaS ingestion platform Fivetran, Airbyte, Google DTS, AWS Glue connectors, or equivalent at the connector-build level, not just configuration

- Strong understanding of OAuth 2.0 flows (auth code, PKCE, client credentials, JWT), rate limiting strategies (token bucket, leaky bucket, per-endpoint quotas), and incremental sync patterns (cursor, watermark, CDC)

- Experience designing shared connector frameworks reusable auth managers, rate governors, state stores not just per-connector scripts

- Ability to author and own TDDs and PRDs that can be handed to a junior engineer with minimal back-and-forth

Nice to have :

- Prior exposure to Databricks Asset Bundles / Declarative Automation Bundles or Lakeflow pipelines

- Experience with the Databricks Python Data Source API (DBR 15.4 LTS+) extremely rare, so treat practical Spark DSv2 Java/Scala background as equivalent

- GCP DTS or Cloud Data Fusion connector experience

- Knowledge of the specific source systems, particularly Social Ads APIs (Meta, LinkedIn, X) or enterprise SaaS (Salesforce, Oracle)