Python Developer — US Software, Data and AI Teams

US employers hire Python developers into three quite different kinds of team, and that distinction matters more than the language does. Backend web teams build and maintain services in Django, Flask or FastAPI. Data and platform teams build the pipelines that move and reshape data. Machine learning teams train, serve and monitor models. All three advertise as "Python developer", and all three pay differently.

What the work involves

On a web backend: HTTP services, database schemas and migrations, authentication, background jobs, and the tests that stop a deploy breaking something. On a data team: ingestion, transformation, scheduling, and the unglamorous work of making a pipeline restartable. On a machine learning team: feature preparation, training, evaluation, and getting a model behind an API where it can be monitored.

Requirements

• Python beyond scripting — data structures, generators, typing, exceptions handled deliberately, and a testing habit
• One framework or toolset in depth for the track you want: Django, Flask or FastAPI for web; pandas, Airflow or dbt for data; PyTorch or scikit-learn for machine learning
• SQL you can be examined on, not just SELECT statements
• Git, code review, and enough Docker and CI to ship your own work
• Cloud exposure — AWS, GCP or Azure — is requested on most postings above entry level

How the pay is structured

• Backend web roles sit against the BLS web developer occupation, a $92,650 median as of May 2025, or the software developer occupation at $135,980 where the work is engineering rather than site building
• Data science roles sit in their own occupation: a $120,230 median, with the lowest tenth under $67,240 and the highest tenth above $199,130
• The domain, not the language, carries the premium. Python is the easiest of these languages to start with, which is exactly why the entry level is crowded and the specialised end is well paid

Before you apply

Read the advert for the occupation, not the title. A posting about endpoints, templates and a CMS is web work. One about experiments, features and model drift is data work. They pay differently and they interview differently, and the title at the top will not tell you which one it is.

Note: pay, remote policy and stack requirements are set by each employer — not by JobGader. Confirm the details on the employer's own advertisement before applying.