Monty is live again: the AI Python Coach for realistic screens (practice out loud, under time pressure). Please leave feedback after your interview so I can keep improving: https://snakesignals.com/interview-prep/
New tool shipped: LinkedIn Analyser (instant notes on recruiter visibility, missing keywords, and Boolean coverage). Check it out: snakesignals.com/linkedin-analyzer/
The signal: “Python + proof” just became the default
A lot of hiring advice is vibes. Here are the numbers that actually describe the market you’re competing in:
GitHub added 36M+ developers in 2025 (about “more than one new developer every second”) and now has 180M+ developers total.
TypeScript finished #1 on GitHub in Aug 2025 with 2,636,006 monthly contributors (+66.6% YoY).
Python is #2 on GitHub and added ~850k contributors in 2025 (+48% YoY), sitting around 2.6M contributors.
1.1M+ public repos now import an LLM SDK (+178% YoY).
~80% of new GitHub developers use Copilot in their first week.
Translation for hiring managers: everyone “uses AI” now. That stopped being differentiating. Evidence of shipping, clarity, and operational judgement is what separates candidates.
Translation for devs: the differentiator is not “LLMs” in your skills list. It’s typed, testable Python that can survive production.
Hiring managers: a 25-minute “proof pack” that beats long interviews
Instead of another 4-stage circus, ask for three artifacts before the live technical:
One PR or diff (real or anonymised) that shows how they change production code
One test they wrote (unit or integration) and why that test exists
One ops note: “what I measured, what I changed, what it cost” (latency, spend, error rate, retries, queue depth, anything real)
Score it on five yes/no checks:
Typing and contracts: do they define inputs/outputs cleanly (Pydantic/dataclasses, consistent error shapes)
Testing intent: do they test behaviour, not implementation trivia
Reliability instincts: retries, timeouts, idempotency, logging, safe defaults
AI judgement: where AI helped, what they measured, and what guardrails exist
Written clarity: can they explain trade-offs without a TED Talk
Why this matters in 2026: time-to-hire is creeping up again, and slow loops lose the best candidates. Indeed’s analysis shows time-to-hire fell hard post-2020, then climbed back toward 2019 levels. Their regression also estimates a 1% rise in quits rate = ~2.7 days faster time-to-hire, while a 1% rise in labour force participation = ~3.1 days slower. Your process speed is a competitive advantage, not an HR metric.
Python ecosystem pulse: the “shipping stack” is obvious in the download data
If you want a practical view of what teams are actually running, PyPI download volumes are a cheat code:
pydantic: ~567M downloads last month
pandas: ~489M downloads last month
fastapi: ~224M downloads last month
ruff: ~125M downloads last month
uv: ~86.7M downloads last month
django: ~32.5M downloads last month
polars: ~38.2M downloads last month
duckdb: ~25.7M downloads last month
Two takeaways:
Python is still the centre of gravity, but typed contracts and fast tooling (Pydantic, Ruff, uv) are basically mandatory now.
Data work is still dominated by pandas in raw volume, but Polars and DuckDB are very real in modern stacks.
LinkedIn Analyser: fix the parts that actually move numbers
LinkedIn has been annoyingly clear about what drives visibility:
Adding a profile photo can drive 21x more profile views and 9x more connection requests
Adding a current position can drive 8x more profile views
Adding your industry can unlock up to 9x more profile views
If you’re early-career, adding education can drive 17x more recruiter messages
Run your profile through the analyzer, then fix the “top third” first:
headline, current role, tech stack keywords, impact metrics, and 2 proof links (GitHub, case study, talk, shipped product).
Job of the week: Lead Python Developer, regulated fintech, London
A UK wealthtech building an all-in-one platform for financial advisers: onboarding, investment accounts, and automated operational workflows. This is production software touching trading and time-critical processes in an FCA-regulated environment.
Comp: up to £130k base + bonus + 6% pension
Setup: hybrid, 3 days onsite (Mon, Tue, Thu)
Must have: Python, Pandas, FastAPI or Django, SQL, strong communication
Scope: lead a backend pod across trading, market data, start-of-day processes; stay 50%+ hands on; mentor 2–3 engineers; own quality and delivery
Process: intro, practical take-home mirroring the work, technical review, architecture discussion, onsite final
If this edition helped, forward it to someone who builds or hires in Python.
Hiring? Contact
Josh Smith
LinkedIn: https://www.linkedin.com/in/python-recruitment/
Email: [email protected]
Phone: 01727 225 552