Python development
Backends, data pipelines and AI agents in Python, on AWS or Azure. Every Python hire is interviewed by a Python specialist rather than a generalist.
What we build in Python
APIs and services
HTTP APIs and background workers with typed request and response models, dependency injection, structured logging and health checks, deployed as containers or serverless functions.
Data work
Ingestion, transformation and reporting pipelines over SQL and document stores, written so a failed run is visible and re-runnable rather than silently skipped.
AI agents and retrieval
Retrieval-augmented generation over your own documents, embeddings in a vector database, and agent workflows with the tool calls and costs logged.
Automation
Scheduled jobs and integrations that replace manual spreadsheet steps, with alerting when a run fails instead of a silent gap in the data.
How we hire for it
Python is not a side skill here. Our vacancy history includes a Python engineer with AWS and AI-agent experience, a Python engineer with vector-database experience, and two dedicated Python technical interviewers, one of them specifically for AI and machine-learning work. Candidates are screened against a written question set and interviewed by someone who writes the language daily.
The same engineering rules as our .NET work
- Typed and linted. Type hints checked in continuous integration, formatting and lint rules enforced, zero warnings at the end of a change.
- Tests with every feature, and a regression test for the class of any bug we fix, not just the one reported instance.
- Errors are never swallowed. Every failure path is logged, surfaced and alertable; a silent retry that hides a broken integration is treated as a defect.
- Secrets from a vault, never from a file in the repository, and every external input validated at the boundary.
Working across stacks
Most of our clients run a mixed estate. A Python service that has to talk to a .NET API, an Angular front end and an Azure SQL database is a normal week here, because all of those sit inside the same team rather than across three vendors.
Have Python work that needs owners?
Send the repository or a description of the pipeline. You get a written plan and an estimate within one business day.