Databricks

The firm's published FAQ on that page states that internships are 12 or 16 weeks long; that early-career hires are placed across seven hub offices – San Francisco, Mountain View, Bellevue, Amsterdam, Berlin, Belgrade, and Bangalore; and that interns are paired with both a manager and a mentor while new grads get a manager plus a six-month onboarding buddy. The same hub names Software Engineering, Data Science, and Product Management as the three role families it publishes prep guidance for. A cohort lands directly into one of these offices and onto a team's roadmap rather than into a rotation, so the application choice is effectively a choice of office and role family – and the firm's new-grad listings ask candidates to apply to a single preferred location and consider only one application per candidate every 30 days.

Three things on Databricks' early-careers hub and live new-grad postings are factual and unusual for a US-headquartered AI infrastructure firm:

• The firm was founded by the original creators of Apache Spark, Delta Lake, and MLflow – the open-source projects underneath a large share of today's data platforms – and the new-grad listings state this directly.
• Early-career hiring is run across seven named hub offices including Belgrade and Bangalore, not only the Bay Area and Seattle.
• Each intern is paired with a manager and a separate technical mentor; each new grad gets an onboarding buddy for the first six months in addition to their manager.

For an applicant picking a small number of technical-platform employers to target, these are the things that separate a Databricks application from one to a generalist big-tech or cloud vendor.