Recomputed every morning · 2026-09-04
What tools do these jobs actually ask for?
Counted from 476 postings whose requirements this board has parsed. Python leads at 254, and the interesting part is how thin the eval-specific tail is.
Named in the posting
Only tools appearing in five or more postings are listed. The pay column needs five published bands before a median is shown, and blank means the sample is too small to say.
| Tool | Postings | Share | Median pay | Bands |
|---|---|---|---|---|
| Python | 254 | 53% | $300k | 104 |
| PyTorch | 110 | 23% | $303k | 46 |
| SQL | 72 | 15% | $260k | 28 |
| Kubernetes | 52 | 11% | $300k | 27 |
| JAX | 42 | 9% | $310k | 17 |
| TensorFlow | 37 | 8% | $381k | 8 |
| Databricks | 33 | 7% | $300k | 17 |
| TypeScript | 30 | 6% | $238k | 19 |
| Rust | 29 | 6% | $310k | 16 |
| Spark | 28 | 6% | $310k | 9 |
| vLLM | 20 | 4% | $300k | 15 |
| Inspect | 11 | 2% | $286k | 5 |
| LangSmith | 8 | 2% | $208k | 7 |
| Label Studio | 7 | 1% | $86k | 6 |
| Snowflake | 5 | 1% | — | 4 |
The spread is about the work, not the tool
Postings naming TensorFlow carry a median midpoint of $381k. Postings naming Label Studio sit at $86k. Nobody is paid more for knowing one library over another. What the gap actually separates is model and infrastructure work from annotation and labeling operations, and those are different jobs that happen to share a board.
Method. Counted from the requirements text of 476 postings, which is 75% of the 637 live roles; the rest did not carry enough description text to parse. A tool counts once per posting that names it. Pay figures are the median midpoint of the published annual USD bands among postings naming that tool, shown only where five or more such bands exist. This list is a fixed vocabulary rather than open extraction, so a tool absent from it is unmeasured rather than unmentioned.
What changed on this board this weekWhat changed since last Thursday: new roles, roles that closed, and pay moves.
Free. One email on Thursdays.