NVIDIA · nvidia.wd5.myworkdayjobs.com · checked today
Software Engineer, Technical Lead – Performance Data Platform
NVIDIA has been transforming computer graphics, PC gaming, and accelerated computing for more than 25 years. It’s a unique legacy of innovation that’s fueled by great technology—and amazing people. Today, we’re tapping into the unlimited potential of AI to define the next era of computing.
Skills, with evidence
- Python
Proficiency with C++, Go, Python.
must have - Data modelling
A track record of crafting APIs and data models that evolved without breaking existing users or deployments.
must have - Machine learning
Practical knowledge of GPU and machine learning workload performance: you can review a performance diagnosis, question its evidence, and decide which signals matter most.
must have · not practised here - SQL
Working knowledge of SQL and TypeScript.
- Airflow / orchestration
Hands-on ability to build and deploy services and tooling yourself, including when using automated development tools, and to verify the result independently through tests and measurements.
- Cost & performance
Describe what queries they may perform, how their results relate to evidence, and the constraints on their cost and response time.
- Governance & security
Identity, access control and multi-tenant build for products that hold customers’ data.
not practised here
Your plan
- The SQL screen: correct, then fast≈ 2 h
SQL
- Python: the data-wrangling round≈ 4 h
Python
- Data modelling: the round most people fail≈ 4 h
Data modelling
- Pipeline design: safe to run twice≈ 4 h
Airflow / orchestration
- Fourteen months of wrong numbersAdvanced
- You fixed it. Who already used it?Advanced
- The payload changed overnightAdvanced
- Creator payout runAdvanced
- Delete me from everywhereAdvanced
- Spark: read the plan Spark actually ran≈ 2 h
Cost & performance
- applyInPandas per country: what Spark ships to Python, and the native rewriteAdvanced
- Filters you wrote in the wrong place: where Catalyst moves themAdvanced
- snappy, gzip or zstd: measure the Parquet codec trade-off yourselfAdvanced
- Find the four suspects in a slow pipeline (one of them is innocent)Advanced
- A filter after a window function: pushed down or not?Advanced
- Say it out loud≈ 1 h
25 drills · Advanced + Intermediate≈ 17 hours
Not covered by the plan: Machine learning, Governance & security.
Readiness
Counted from drills you have completed anywhere on D8LooP.
leaves in 4 dremoved the moment NVIDIA closes it