Market · Machine learning
Peak Python
220 of the 351 machine learning jobs we read name Python: 63%, the highest Python share of the five pages.
The jobs are a mountain range. Each peak is a skill the page charts, and its height is the share of the 351 jobs that name it. Swing the arm on the table to read a peak.
By the Glassbox team · the week of 12 October 2026 · a four minute read
This story keeps to one week, 12 October 2026. The machine learning page starts a new count every Monday.
| Skill | Ridge | Of 351 jobs | Share | Employers | Of 194 senior jobs |
|---|---|---|---|---|---|
| Python | Languages | 220 of 351 | 63% | 128 | 64% |
| C++ | Languages | 54 of 351 | 15% | 32 | 13% |
| SQL | Languages | 41 of 351 | 12% | 31 | 11% |
| Go | Languages | 20 of 351 | 6% | 14 | 8% |
| Machine learning | AI | share only | 92% | ||
| LLMs | AI | 158 of 351 | 45% | ||
| PyTorch | AI | 150 of 351 | 43% | 99 | 74 of 194 · 38% |
| ETL / data pipelines | Data | share only | 21% | ||
| Statistics | Data | share only | 21% | ||
| Apache Spark | Data | share only | 12% | ||
| Kubernetes | DevOps | share only | 13% | ||
| CI/CD | DevOps | share only | 11% | ||
| Observability | DevOps | share only | 10% | ||
| AWS | Cloud | 63 of 351 | 18% | 45 of 194 · 23% | |
| Google Cloud | Cloud | share only | 12% | ||
| Azure | Cloud | share only | 7% | ||
| MLOps | More skills | 84 of 351 | 24% | 57 of 194 · 29% | |
| Reinforcement learning | More skills | 58 of 351 | 17% | 24 of 194 · 12% | |
| Feature engineering | More skills | 28 of 351 | 8% | 25 of 194 · 13% |
Every skill the machine learning page charts, to one scale on one baseline. A peak’s height is the share of the 351 jobs that name it. The pale peaks behind are the share in the 194 senior machine learning jobs, for the nine skills the page compares. A job can name skills on several ridges.
Source: Glassbox count, week of 12 October 2026 · how we count
Four languages, one tall peak
Python is named in 220 of the 351 jobs, 63%, by jobs at 128 of the 174 employers. The next language peak is C++, 54 jobs. Then SQL, 41. Then Go, 20. Only four languages are named by jobs at 10 or more employers, so only four are drawn.
232 of the 351 jobs name at least one language. 115 name two or more. The other 119 name none (351 − 232 = 119).
Python’s peak on each of the five pages we count, to one scale. Machine learning: 220 of 351 jobs. Entry level software: 276 of 559. AI: 258 of 567. DevOps: 246 of 567. Software engineering: 1,092 of 2,757.
Source: Glassbox count, week of 12 October 2026 · how we count
The ridge almost every job stands on
343 of the 351 jobs name a skill in AI. The tallest peak is the range’s own name. This page counts machine learning jobs, and 92% of them name machine learning among their skills. So 8% don’t (100 − 92 = 8).
Beside it, LLMs: 158 of 351 jobs, 45%. And PyTorch: 150 of 351, 43%, named by jobs at 99 employers.
The lower ridges
Data, 172 jobs: ETL / data pipelines 21%, statistics 21%, Apache Spark 12%. DevOps, 102 jobs: Kubernetes 13%, CI/CD 11%, observability 10%. Cloud, 73 jobs: AWS 18%, Google Cloud 12%, Azure 7%.
The page gives most of these as shares of the 351 jobs, so those peaks carry a share and no count.
The far range
Behind nine of the peaks stands a paler one: the same skill in the 194 senior machine learning jobs. PyTorch is 43% here and 38% there. Reinforcement learning, 17% and 12%. MLOps goes the other way, 24% here and 29% there. AWS, 18% and 23%. Feature engineering, 8% and 13%. Python barely moves: 63% and 64%.
The visitors’ book
What the 351 jobs say about themselves, the way a hut keeps a book.
Years asked
176 of the 351 jobs state a number of years. 175 don’t.
- 0 to 2 · 32
- 3 to 4 · 37
- 5 to 7 · 75
- 8 or more · 32
- no number · 175
Level
Senior means a senior, lead or principal level.
- senior · 194
- junior · 50
- mid · 27
- not stated · 80
Where
The place each job names.
- US · 179
- elsewhere · 113
- no country we can read · 59
- Remote
- 130 of 351 · 37%
- Internships
- 27 of 351 · 8%
- Employers
- 174 · Reddit has the most, 15 of the 351
Each band’s width is its count of the 351 jobs. Hatched: the job states no number, no level or no country we can read. Lime: the most common ask, 5 to 7 years.
Source: Glassbox count, week of 12 October 2026 · how we count
Seen from here
Five of the 351, titles and employers only, as the page lists them.
- Research Scientist - 3D Reconstruction (SfM & SLAM)Spaitial
- Staff Machine Learning Engineer, Ads Foundational RepresentationsReddit · asks 7 years
- Research Engineer - Eval PlatformMistral.ai · asks 4 years
- Machine Learning Engineer - Voice AI & Generative Music (Part-time)MWDN · asks 3 years
- Credit Model Validation Manager (Machine Learning & NPV Models)Monzo
Curious how many of these skills are already on your CV?
One week, 12 October 2026. The machine learning page counts again on Monday.
How we counted
This page counts the machine learning jobs: titles that name machine learning, deep learning, computer vision, NLP, or applied or research science.
The job type comes from the title. Senior means a senior, lead or principal level in the same job type.
We look for about 250 named skills, by fixed rules, in the 351 ads long enough to read. The weekly count is published every Monday.
The range: every skill the page charts is a peak, drawn to one scale on one baseline. Where the page gives a count we print it. Where it gives only a share we print only the share. The pale peaks are the page’s figures for the 194 senior machine learning jobs. Python’s share on the other four pages comes from each page’s languages table.
The jobs come from public job boards, read every six hours: employers' own boards on Greenhouse, Ashby, Lever and Workable, and job sites such as Remote OK, Remotive, Arbeitnow, Jobicy, Himalayas and We Work Remotely. Each job counts once, however many boards or cities list it.
What we can't see: jobs on boards we don't read, and jobs never posted in public; skills a job names in words that aren't on our list; whether a job is filled, how many people apply, or who gets hired.