Working with Python?

Python scripts nobody owns any more? We take them over.

Most Python we inherit started as one script for one job, grew for a few years, and then the person who understood it moved on. We read what is actually there, get it running on a current version, and leave it in a state the next person can pick up.

  • Since 2015Fixing technology
  • FreeEstimates
  • US teamReal engineers

cron.log

still failing

What it is

Python at a glance

Python, created by Guido van Rossum in 1991, is a general-purpose high-level language. It is read like plain English by people who have never seen it before, which is one reason it ended up running so much of the world's glue code, reports, and overnight jobs.

It dominates data science, machine learning, web back ends built on Django, Flask or FastAPI, automation, scientific computing, and DevOps scripting. Python 2 reached end of life January 1, 2020, so Python 3 is the only supported line. The language is looked after by the Python Software Foundation and a large contributor community.

Type
General-purpose programming language
First released
1991
License
Python Software Foundation License
Maintained by
Python Software Foundation and contributors
Typically used for
Back ends, data pipelines, machine learning, automation, scripting

The honest version

Python pros and cons

No technology is all upside. This is where Python earns its place and where it tends to cause trouble.

Where Python is strong

  • You can actually read itPython code tends to look like a description of what it does, which matters a lot when the person who wrote it is no longer around to explain.
  • An enormous ecosystemNumPy, Pandas, Django, Flask, FastAPI, Celery, Scikit-learn and PyTorch together cover most of the problems a business ever needs to solve with code.
  • It is already on the serverPython ships with almost every Linux distribution and every developer laptop, so there is rarely a fight to get a script running somewhere new.
  • Great at the boring workFile shuffling, report generation, scraping, API glue and overnight jobs are where Python quietly saves the most time.
  • Strong in data and MLIf a question starts with a spreadsheet and ends with a model, the shortest path between the two is almost always Python.
  • Modern async and type hintsRecent versions give you real async I/O and optional type checking, so new code can be both fast and safer to change.

Where Python bites

  • Python 2 still haunts old estatesThe split from Python 2 to Python 3 was painful through the 2010s, and plenty of internal scripts still quietly assume Python 2 behaviour that no modern server provides.
  • Each release deprecates somethingEvery minor version removes or renames something people relied on, such as collections.Mapping, distutils or assumed dict ordering, so unmaintained code rots on a predictable schedule.
  • Packaging is a crowded fieldVirtual environments come in several flavours, and venv, pipenv, poetry, uv and conda all solve overlapping pieces of the same problem in different ways.
  • The GIL limits CPU-bound threadsFor number crunching across many cores, the global interpreter lock means real parallelism needs processes, native extensions or a different approach entirely.
  • Binary dependencies pin badlyLibraries that wrap C or Rust code, such as lxml or older NumPy, can refuse to build on a new OS and are painful to reproduce years later.

Sound familiar?

Common Python problems we fix

The Python problems that bring people to us, and what you have once each one is fixed.

print statement · Python 2.7

An old Python 2 script that nobody will touch

A business-critical script written against Python 2, running on a server that is years overdue for an upgrade, with no tests and no original author.

Running on Python 3, with tests

author: unknown

The person who wrote it left

A folder of scripts with no README, no requirements file, and a vague memory that one of them emails the invoices on the first of the month.

Documented, owned, repeatable

/usr/bin/python: No such file or directory

Cron jobs broken after a server upgrade

A new OS release removed the old Python binary or moved it, and overnight jobs have been silently failing for weeks before anyone noticed.

Jobs running, with alerts if they don't

Failed to build wheel

pip install refuses to work

Pinned versions that no longer build on a current OS, or private packages that depend on libraries that have since been renamed or removed.

Reproducible environment, pinned properly

runs for 6 hours

A report that takes all night

A Pandas or pure-Python job that grew past the size it was designed for and now overlaps with the next day's run.

Finishes in minutes, same answers

MemoryError

It falls over on bigger files

A script that loads a whole file into memory at once, fine for last year's data, fatal for this year's.

Streams the data, finishes cleanly

What we do

Python help, from rescue to upkeep

Call us for one broken script, or hand us the whole folder.

Rescue
A script that stopped working after an upgrade, or a job that was never reliable to begin with. We get it running, then find out why it stopped.
Takeovers
Inheriting Python written by somebody who is no longer around, reading it properly, and giving you back something you can hand to anyone.
Python 2 to 3 upgrades
Moving long-lived scripts onto a supported runtime, with tests so the output after the move matches the output before it.
Web back ends
Django, Flask and FastAPI work: new features, upgrades between major versions, and performance work on slow endpoints.
Data and automation
Pipelines, scheduled jobs, API integrations and reports that have to be right every time.
Managed Tech
Monitoring, dependency updates and a team to call, so scripts stop quietly rotting between hands.

How it works

From first message to fixed

  1. Look Free estimate

    Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.

  2. Fix

    We fix the cause, not the symptom. If something is down, getting it back up comes first.

  3. Harden

    Backups, monitoring, updates and access, so the same problem can't quietly come back.

  4. Hand back or stay on

    You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.

Often found together

Running Python alongside something else?

We work on the whole stack, so the problem can be wherever it actually is.

Questions

Python questions

Something else? Call+1 737 727 7327

Can you upgrade our old Python 2 scripts to Python 3?

Yes. We read the code, capture what it currently produces so we have something to compare against, move it to Python 3 piece by piece, and run the old and new side by side until the outputs match. Scripts with no tests get tests as part of the job.

Our Python script was built by someone who left. Can you take it over?

That is most of what we do with Python. We read what is actually there, write down how it is meant to work, get it running cleanly on a current Python with a pinned environment, and leave it documented enough that the next person has a chance.

Why do my Python cron jobs break after a server upgrade?

Usually the OS has moved or removed the Python binary the cron line was pointing at, or a library the script depended on no longer builds, or a feature the script used was deprecated in the newer Python. Pinning the environment and using a virtual env fixes this for good.

Should we rewrite our Python script as something faster?

Rarely the right first step. Python slowness is almost always a loop doing too much work, a Pandas call fighting itself, or a database query that needs an index. We measure first, and a rewrite only comes up if the measurements justify it.

Why does pip install fail on this project?

Most often an old requirements file pins versions that no longer build on a current OS, or depends on a package that has since been renamed. We rebuild the environment against versions that actually install, keep the behaviour the same, and lock it so it keeps working.

How much does Python help cost?

Estimates are free. Send us the script or describe the job and you get a price before any work starts.

Free estimate

Python script giving you trouble?

Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.

  • No charge to find out what it will take
  • If it's urgent, say so

Replies come from a person. By sending, you authorize us to email you back. No marketing.