CodeCut is a publication that tests and compares Python, data science and AI tools. Comparisons run each tool on the same input, on named hardware, at listed versions, and the notebook is published so readers can rerun the test themselves. Khuyen Tran founded CodeCut in 2024. The publisher is CodeCut Technologies LLC.
What CodeCut publishes
Tool comparisons that answer which tool to use for a given job, with measured results. Recurring subjects are dataframe and query engines, document parsing and OCR, LLM application frameworks, agent tooling, and the Python packaging and testing stack. Current comparisons are listed at codecut.ai/blog.
Deep Dives, sent every Tuesday: long-form tutorials that work through a single tool or workflow from start to finish, with code you can run as you read. Example: VCR.py: Make Public API Tests Repeatable in Python
Quick Tips, sent every Thursday: one Python or AI tool or feature per issue, shown in a short code example and readable in a few minutes. Example: Swap AI Prompts Instantly with MLflow Prompt Registry
Code for the articles is at github.com/khuyentran1401/codecut-blog, and most articles link their notebook directly.
Who CodeCut is for
CodeCut is written for people who already work in Python and have to pick between tools: data scientists, analytics and data engineers, and engineers building LLM applications. Articles assume you can read code, and skip the introduction to Python itself.
Who writes CodeCut
Khuyen Tran writes CodeCut, with invited contributors who are credited on their own work.
Before CodeCut, she was an MLOps Engineer and a Senior Data Engineer at Accenture, building data systems for enterprise clients. She has written more than 180 articles as a top writer on Towards Data Science, wrote Production-Ready Data Science and Efficient Python Tricks and Tools for Data Scientists, and has talked about this work on the Talk Python To Me podcast. Her code is on GitHub, CodeCut’s repositories are at github.com/CodeCutTech, and she posts on LinkedIn and X.
Editorial and sponsorship policy
Some articles are sponsored, and those say so in the first line. Sponsors do not review or approve conclusions: CodeCut runs the tests and reports what they show, including results that are unfavorable to the sponsor.
Companies interested in reaching this audience can read the terms on the sponsor page.
Contact
Email khuyentran@codecut.ai.