Remote Python Engineer: Codebase Explorer for AI Reasoning

Obsidian; New York (NY); On-site; Full time

Job summary

Mercor is seeking experienced Python engineers to build the dataset that teaches AI agents to reason about real codebases. This track covers production Python codebases such as Django, Twisted, Scrapy, SymPy, mypy, pip, coverage.py, mitmproxy, Certbot, Tornado, Werkzeug, sqlfluff and Meson.

Contract work pays at $130 per approved task and is performed remotely in Mercor Studio, with fully remote and asynchronous collaboration across US, India, and Nigeria time zones.

Qualifications

• 3+ years of professional software engineering experience with Python.
• Ability to navigate large, unfamiliar Python codebases and explain runtime behavior.
• Fluency in Python internals: import machinery, descriptors and metaclasses, async and event loops, packaging.
• Experience with web frameworks, tooling, type checkers, or scientific libraries; focus on architecture and system design.
• Excellent written English; documentation and rubrics must be precise.
• Comfortable with git at commit level, containerized environments, and command-line tooling.

Responsibilities

• Select tasks from a pre-validated pool of Python repositories with specific commits and rubrics.
• Explore the repository until you understand the subsystem and relevant history.
• Rewrite the task's question to challenge frontier coding agents while staying well-defined.
• Adjust rubrics and foils to reward correct reasoning and reject plausible-but-wrong answers.
• Run the full validation loop in Studio — Check, Golden, Run, Analyze — and iterate.
• Document work in the task README and address reviewer feedback.

Skills

Python

Codebase navigation

Runtime reasoning

Git & CI

Containerized environments

English communication

Tools

Django

Twisted

Scrapy

SymPy

mypy

pip

coverage.py

mitmproxy

Certbot

Tornado

Werkzeug

sqlfluff

Meson

Job description

Mercor is seeking experienced Python engineers to build the dataset that teaches AI agents to reason about real codebases. This track covers production Python codebases such as Django, Twisted, Scrapy, SymPy, mypy, pip, coverage.py, mitmproxy, Certbot, Tornado, Werkzeug, sqlfluff and Meson.

Contract work pays at $130 per approved task and is performed remotely in Mercor Studio, with fully remote and asynchronous collaboration across US, India, and Nigeria time zones.