Anuj Ayav worked on the ABrain-One/nn-dataset repository, where he developed a code similarity analytics module designed to efficiently detect similar code segments across neural network models. Leveraging Python, he implemented MinHash and Locality-Sensitive Hashing (LSH) techniques to reduce the computational cost of code comparison from quadratic to near-linear, enabling scalable analysis within large datasets. His work focused on algorithm design and data analysis, integrating the new module with existing dataset tooling to support future analytics. The solution improved the workflow for assessing model and code reuse, laying a technical foundation for more advanced similarity metrics in machine learning pipelines.