Demo Portfolios
Explore live portfolios built from real data — publications, repositories, dependency graphs, and community health metrics.
NumFOCUS Sponsored Projects
Projects sponsored by the NumFOCUS foundation — the backbone of open scientific computing.
Explore portfolio →SciLifeLab Publications
Research publications from SciLifeLab (Science for Life Laboratory), Sweden's national research infrastructure for molecular biosciences — a collaboration between Karolinska Institutet, KTH, Stockholm University, and Uppsala University.
Explore portfolio →Top PyPI Packages
The most downloaded Python packages on PyPI, mapped to their GitHub repositories. This portfolio tracks the most critical packages in the Python ecosystem by download volume, spanning: - Cloud infrastructure (boto3, botocore, s3transfer) - Web and networking (requests, urllib3, certifi) - Data science and machine learning (numpy, pandas, scikit-learn, tensorflow) - Cryptography and security (cryptography, cffi, pyOpenSSL) - Developer tooling (setuptools, pip, wheel, packaging) - Core utilities (typing-extensions, six, python-dateutil) Together these packages form the backbone of Python's digital infrastructure, depended upon by millions of projects worldwide.
Explore portfolio →Fundamental AI
The frameworks, tools, and research powering artificial intelligence. This portfolio maps the full AI stack — from foundational research papers to production infrastructure: - Deep learning frameworks (PyTorch, TensorFlow, JAX) - Large language models and inference (Transformers, vLLM, Ollama, llama.cpp) - Computer vision (OpenCV, YOLO, Stable Diffusion) - NLP and speech (spaCy, Whisper, LangChain) - Reinforcement learning (Gymnasium, Stable Baselines) - MLOps and experiment tracking (MLflow, Weights & Biases, Ray) - Seminal research (Attention Is All You Need, ResNet, BERT, GPT, and thousands more) PRSM reveals what's underneath: the hidden dependencies, the infrastructure nobody credits, the actual foundations of AI.
Explore portfolio →CZI EOSS Grantees
Projects funded by the Chan Zuckerberg Initiative's Essential Open Source Software for Science (EOSS) program. CZI EOSS supports software projects that are essential to biomedical research and the broader scientific community. The program recognizes that much of modern science depends on open source software, and invests in the maintenance, growth, and sustainability of these critical tools. This portfolio spans a wide range of scientific domains: - Bioinformatics & genomics (GATK, samtools, Nextflow, Snakemake, Bioconductor) - Neuroimaging (fMRIPrep, FreeSurfer, MNE-Python, Nilearn) - Microscopy & imaging (napari, ImageJ/Fiji, QuPath, Bio-Formats) - Single-cell analysis (Scanpy, Seurat, scvi-tools, AnnData) - Core scientific computing (NumPy, SciPy, pandas, Matplotlib, scikit-learn) - Statistical modeling (Stan, PyMC, ArviZ) - Data infrastructure (JupyterLab, Dask, conda-forge, Zarr) Together these ~200 projects form the backbone of open source scientific software.
Explore portfolio →Bioconductor Packages
The complete Bioconductor software ecosystem for bioinformatics in R. Bioconductor provides tools for the analysis and comprehension of high-throughput genomic data. Founded in 2001, it has grown to ~3,700 packages spanning every major area of computational biology: - Gene expression analysis (DESeq2, edgeR, limma) - Single-cell analysis (SingleCellExperiment, scran, scater) - Proteomics and mass spectrometry (MSstats, OpenMS, mzR) - Flow cytometry (flowCore, flowClust) - Genomic annotation (GenomicFeatures, AnnotationDbi, biomaRt) - Sequence analysis (Biostrings, GenomicAlignments, Rsamtools) - Epigenomics (methylKit, ChIPseeker, bsseq) - Visualization (Gviz, ggbio, ComplexHeatmap) This portfolio maps each package to its upstream GitHub repository where available, falling back to the official Bioconductor mirror (github.com/bioc/*) otherwise — giving a more complete view of the authors, contributors, and development communities behind the Bioconductor ecosystem.
Explore portfolio →Citation-File Projects (2026-05-08)
Every GitHub repository known to publish a CITATION.cff or CITATION.bib file as of 2026-05-08. The Citation File Format (CFF) is a community standard that lets software authors declare exactly how they want their work to be cited — author lists, DOIs, version, release date, related preferred citations. A repository that ships a CITATION file is making an explicit, machine-readable statement: this is research software, and here is how to credit it. This portfolio is a snapshot of that population: roughly 47,000 repositories spanning the long tail of scientific computing — niche domain tools, dissertation code, lab software, individual researcher libraries, and the well-known frameworks alongside them. Together they offer a view of research software practice that does not depend on package-registry presence, citation in a journal, or membership in a foundation: it captures everyone who took the time to say "please cite me this way." Built from the latest citation-file index, normalized to one canonical GitHub URL per repository.
Explore portfolio →