CSE Student at IIIT Vadodara, Gujarat, India 🇮🇳
I'm interested in backend engineering, healthcare data systems, and open source development.
Currently contributing to KathiraveluLab/Diomede for GSoC 2026. Building a pipeline for creating shareable albums from locally stored DICOM medical images (Project #13), with a focus on consuming Niffler's existing metadata extraction pipeline rather than reimplementing it.
| PR | Description | Tests |
|---|---|---|
| #132 | Niffler-consuming album indexer — POST /index-from-niffler, generator-based CSV streaming, memory-efficient batch DB queries, path traversal protection (CWE-22) |
18 |
- Reads Niffler's CSV output and feeds it into the album index pipeline
- Generator-based processing for memory-efficient handling of large DICOM datasets (10,000×10,000 scale)
- Batched SQLite queries using
INoperator to avoid loading all paths into memory - Path traversal security using
os.path.normpath+os.path.commonpath - Flask REST endpoint that integrates cleanly with the existing
DICOMAlbumCreatorpipeline
- Languages: Python, C, C++
- Frameworks: Flask, pandas, SQLAlchemy, numpy, Node.js, Express.js
- Tools: pydicom, pytest, Git, SQLite
- Domains: Backend APIs, Medical Imaging (DICOM), Open Source
- GitHub: @aasthathakkar
- Location: Gandhinagar, Gujarat, India (IST — UTC+5:30)