Northwind Analytics started as a way for a handful of data engineers to keep working the way we liked: short engagements, direct client contact, and no bench of people waiting to be staffed. We've kept it that way.
We work with mid-market companies — usually somewhere between "the founder built the first dashboard in a spreadsheet" and "we have a real data team but no time to fix the pipeline underneath it." Most of our work is unglamorous: migrating a database, rebuilding an ETL job that one person understood, putting a proper staging environment in front of a production database that never had one.
We're a distributed team with people across Europe and North America, loosely anchored around Zurich. No office, no floor plan, no ping-pong table photos.
Engagements are typically one senior engineer plus one generalist, working directly inside your existing stack rather than standing up a parallel one. We favor boring, well-understood technology — managed Postgres over anything exotic, plain Python over frameworks-of-the-month — because boring is what's still running in two years.