RMRui Mendes Portfolio
← Back to My work

Farfetch · 2022–2024

DataHub — Ingestion as a Service

A self-service data ingestion capability that reduced dependency on manual workflows and gave teams more autonomy over data pipeline operations.

Context & approach

At Farfetch, data dependencies moving into the analytics and ML ecosystem previously passed through a central team and a manual configuration queue. The product challenge was ownership, visibility, repeatability and speed.

The product evolved into Ingestion as a Service inside DataHub, enabling teams to configure, deploy, modify and decommission ingestion pipelines with governance considerations.

The rollout moved from manual scripts to DataHub-supported flows, then formal service requests and SLAs, and finally towards full self-service for new ingestions.

Outcomes

  • Reduced time-to-market for ingestion from 32 days to a 1–2 day target.
  • Saved an estimated 85%+ manual effort.
  • Created a scalable internal platform model across three engineering teams.