Python Skill

Python Expert

Overview

Advanced expertise in Python development — from scripting and automation to production-grade APIs, data engineering, and AI/ML systems. Specialized in idiomatic, performant, and maintainable Python code.

1. Core Python

• Python 3.10+ features (match/case, type unions, | syntax)
• Type annotations & typing module (TypeVar, Protocol, Annotated, TypedDict)
• Dataclasses, __slots__, __init_subclass__, descriptors
• Context managers (contextlib, __enter__/__exit__)
• Generators, iterators, yield from, send()
• Decorators (functools, class-based, parametrized)
• Metaclasses and __class_getitem__
• Abstract base classes (abc.ABC, abstractmethod)
• __dunder__ methods and operator overloading

2. Async & Concurrency

• asyncio event loop, tasks, gather, wait, timeout
• async/await, async generators, async context managers
• aiohttp, httpx for async HTTP
• ThreadPoolExecutor & ProcessPoolExecutor
• concurrent.futures, multiprocessing, threading
• GIL awareness and CPU-bound vs I/O-bound strategies
• anyio / trio compatible patterns

3. Web & API Development

• FastAPI: routers, dependency injection, middleware, lifespan, OpenAPI
• Django: ORM, signals, custom managers, class-based views, DRF
• Flask: blueprints, extensions, application factories
• Pydantic v2: validators, model_validator, computed fields, serialization
• OAuth2 / JWT authentication patterns
• WebSockets with FastAPI / Starlette
• Rate limiting, CORS, request validation

4. Data & Databases

• SQLAlchemy 2.x (async sessions, mapped dataclasses, relationships)
• Alembic migrations
• PostgreSQL (asyncpg, psycopg3), SQLite, MongoDB (motor)
• Redis (redis-py, aioredis) — caching, pub/sub, queues
• Pandas, Polars for data wrangling
• Arrow, Parquet, CSV/JSON ETL pipelines
• Celery + Redis/RabbitMQ for task queues

5. Testing & Quality

• pytest — fixtures, parametrize, marks, plugins
• pytest-asyncio for async tests
• unittest.mock, MagicMock, AsyncMock, patch
• hypothesis for property-based testing
• coverage.py, branch coverage
• mypy strict mode, pyright type checking
• ruff linting & formatting, black, isort
• pre-commit hooks

6. CLI & Scripting

• argparse, click, typer for CLI tools
• rich for terminal output (tables, progress bars, panels)
• pathlib for file system operations
• shutil, tempfile, os, subprocess
• dotenv / pydantic-settings for configuration
• Shell scripting integration via subprocess.run

7. AI & ML Integration

• OpenAI / Anthropic SDK usage and streaming
• LangChain, LlamaIndex integration patterns
• transformers (HuggingFace) — inference, fine-tuning
• NumPy, SciPy for numerical computing
• scikit-learn pipelines, cross-validation, serialization
• Async LLM client patterns, structured output parsing
• Vector databases (Chroma, Weaviate, Pinecone)

8. Packaging & DevOps

• pyproject.toml, setup.cfg, hatch, poetry, uv
• Virtual environments: venv, conda, uv venv
• Docker multi-stage builds for Python apps
• GitHub Actions CI for Python (lint → test → build → publish)
• PyPI publishing (twine, build)
• Semantic versioning & changelogs

9. Performance & Optimization

• Profiling: cProfile, line_profiler, memray
• functools.lru_cache, cache, cached_property
• __slots__ for memory reduction
• Cython, ctypes, cffi for C extensions
• numpy vectorization over loops
• Lazy evaluation and generator pipelines
• Connection pooling and batch processing

10. Security Best Practices

• Input validation with Pydantic / marshmallow
• SQL injection prevention (parameterized queries)
• Secrets management (keyring, env vars, vaults)
• bandit static analysis
• Dependency auditing (pip-audit, safety)
• Secure file handling and path traversal prevention

Core Competency Summary

• Write idiomatic, type-safe, production-ready Python
• Design async-first API services with FastAPI + SQLAlchemy
• Build robust CLI tools and automation scripts
• Integrate AI/ML models and LLM APIs
• Test, lint, and ship Python packages with modern tooling
• Optimize for performance, security, and maintainability