James writes about code integrity for practicing engineers and helps teams wire Codequiry into their CI and review pipelines.
General
9 min
James Okafor
•
9 hours ago
Line diffs collapse when students rename variables, reorder functions, and extract methods. Here's how token fingerprints and AST normalization catch refactored copies, and how to triage hundreds of similarity reports without drowning in false positives.
General
7 min
James Okafor
•
6 days ago
Web code plagiarism hides in plain sight when students copy from Stack Overflow, GitHub, or tutorials and rename a few variables. This post shows how to teach detection as a skill, design assignments that surface copied web code, and use a source-aware checker like Codequiry to see the evidence.
General
8 min
James Okafor
•
1 week ago
A head-to-head evaluation of AI code detection across ChatGPT-4o, GitHub Copilot, Claude 3.5 Sonnet, and Gemini 1.5 Pro. One pattern kept surfacing: text-only detectors miss refactored LLM code, while structural and multi-signal checks hold up.
General
11 min
James Okafor
•
2 weeks ago
When students rename variables, extract methods, and reorder statements to hide copied code, which detection algorithm actually holds up? A controlled experiment pits winnowing, token-based matching, and AST structural hashing against a ladder of refactoring transformations — and reveals why single-technique checkers miss the cases that academic-integrity panels care about most.
General
9 min
James Okafor
•
3 weeks ago
A study of 5,000 Java assignments from three US universities found that nearly one in four contained code blocks directly traceable to Stack Overflow answers — yet traditional similarity checkers missed them all. We applied token-sequence fingerprinting and a web index of 1.2 million programming snippets to surface hidden web plagiarism at scale.
General
11 min
James Okafor
•
3 weeks ago
Perplexity-based detectors aren’t magic — they measure how surprising a sequence of code tokens would be to a model trained on real human code. This report breaks open the inner math, real false-positive rates from Stanford and Edinburgh benchmarks, and why the strongest detectors stack statistical signals with AST fingerprinting and web-source checks.
General
6 min
James Okafor
•
1 month ago
Coding bootcamps that graduate thousands of developers a year are shifting from manual spot-checks to automated AI code detection on every single submission. Here’s a hands‑on guide to building that pipeline with Codequiry’s API — from single‑file scanning to a full CI‑grade batch checker you can run in a GitHub Action.
General
8 min
James Okafor
•
1 month ago
From the early days of the Unix diff command to the rise of MOSS, JPlag, and AI-powered detectors, code similarity detection has undergone a quiet revolution. This retrospective traces the key technical milestones—tokenization, ASTs, fingerprinting, web-source matching, and the new frontier of AI-generated code—showing how each layer made plagiarism harder to hide. See how modern platforms like Codequiry unify these techniques into a single pipeline.
General
8 min
James Okafor
•
1 month ago
We ran 2,400 real student Java assignments through MOSS, JPlag, Turnitin, and Codequiry with known ground truth. The F1 scores and false positive rates diverged by over 20 percentage points. Here’s what worked, what broke, and how to run your own benchmark.
General
11 min
James Okafor
•
1 month ago
When a mid-sized software firm started scaling external contractor contributions, they needed a way to verify code originality at scale. By integrating Codequiry’s API into their CI/CD pipeline, they built an automated gate that flagged plagiarized open-source snippets and AI-generated code before it ever hit the main branch. Here’s a look at their process, the false-positive tradeoffs, and the integration that saved their legal review hours.
General
9 min
James Okafor
•
1 month ago
When a growing SaaS startup suspected a third‑party deliverable contained unattributed open‑source code, they built a verification pipeline that now scans every contractor commit. Here’s what they learned about trade‑offs, accuracy, and the tools that stuck.
General
8 min
James Okafor
•
1 month ago
Modern CS courses often span multiple programming languages, but traditional plagiarism tools like MOSS only compare files within the same language. This case study walks through a practical detection pipeline that catches students rewriting Java logic in Python, using token-normalised AST comparisons and Codequiry’s cross-language API.