Data Engineer

Most AI and analytics projects fail on the data, not the model. Clients arrive with six systems, three spreadsheets that matter more than anyone will say out loud, and no agreement on what counts as a customer. Your work starts exactly there.

You'd build ingestion and transformation pipelines, define the models the product and the reporting sit on, and set the quality checks that stop nonsense reaching a dashboard someone makes decisions from.

The role sits beside the AI engineers, who need retrieval sources that are current, and the backend engineers who own the operational side.

Your work is the foundation for everyone else's here. Retrieval, agents and evaluations are only as good as the pipelines feeding them.

• Move the data. Pipelines from client systems, including the one with no API and a CSV export button.
• Model the domain. For analytics and retrieval, documenting what each field means in their business.
• Make breakage loud. Quality checks and alerting, so silent corruption doesn't become a board slide.
• Feed the AI work. Clean, current sources for retrieval and evaluation, refreshed on a schedule people can rely on.
• Watch the meter. Cost and runtime, because an unwatched pipeline gets expensive without anyone noticing.

• Three Years of Pipelines in Production: running daily, with someone depending on the output.
• dbt or Equivalent Modeling: tested, documented models instead of a folder of views nobody dares to change.
• Warehouse Depth: BigQuery, Snowflake, Postgres or DuckDB, and the cost model that comes with your choice.
• Streaming Where It Earns It: Kafka or a managed queue, used where batch won't do.
• Data Quality as Code: freshness, row count and schema tests that wake someone before the client spots it.
• Python That Survives Review: typed, tested and packaged, not a notebook renamed to .py.
• Governance Basics: personal data, retention and access decided and written down well before the audit.
• SQL and Python Fluency: plus a warehouse you know well: BigQuery, Snowflake, ClickHouse or similar.
• Orchestration Experience: Airflow, Dagster or whatever kept your jobs in order and your nights quiet.
• Messy Source Survivor: you have decided what to trust in a contradictory dataset, and defended the decision.
• Thinks in Backfills: schema changes, reruns and idempotency occur to you before the first run, not after.
• Explains Number Changes: when a metric moves, you can say why, and be right.
• Fluent English, Written and Spoken: most of our work happens in a written thread or on a call, and both have to be clear. Arabic is a genuine advantage across the region.
• AI-Aware Modeling: you design data with retrieval and evaluation in mind, not only dashboards.

• Skeptical of data: you check before you believe a total.
• Tidy: naming, documentation and lineage aren't chores to you.
• Collaborative: you work with analysts and engineers rather than throwing tables over a wall.

• Streaming: Kafka, Kinesis or change data capture.
• Vector stores and embedding pipelines for retrieval.
• dbt or another transformation tool used somewhere large.

• Remote across MENA. Work from wherever you are in the region, with hours built around your team's overlap rather than someone's attendance spreadsheet.
• Work that ships. Products go live in weeks, not quarters. When someone asks what you have been building, you can open it on your phone and show them.
• Small teams, no ticket queue. Four to eight people carry a product from the first client call to handover. Nobody hands you a spec and disappears.
• Kit and tools you choose. A budget for your setup, and for the editor, licenses and services that make you fast. No three-week procurement saga.
• A learning budget with time attached. Money for the course or the conference, plus room in the week to use it. The second half is the part most places skip.
• Client contact from day one. You sit in the room where the decisions get made. Plenty of engineers wait a decade for that seat.
• Time off people take, and no timesheets. We measure engagements, not humans in fifteen-minute increments. Book the holiday, go, come back.