.. _appfl_integration:

Integration with APPFL
=======================

AIDRIN collaborates closely with the `APPFL framework <https://appfl.ai/en/latest/index.html#>`_ to improve the quality and reliability of datasets used in federated learning workflows. This partnership brings together AIDRIN’s advanced data readiness capabilities and APPFL’s powerful federated learning infrastructure to help researchers and practitioners work with high-quality, AI-ready data from the start.

Through this integration, users of APPFL can benefit from AIDRIN’s tools to assess, interpret, and improve datasets before and during federated learning experiments. This ensures that the data used across distributed systems meets key standards for completeness, balance, and cleanliness—factors that are critical for trustworthy and efficient AI training.

This collaboration reflects a broader effort to address data quality challenges in federated learning by embedding practical, customizable solutions into real-world ML workflows. The integration enables users to go beyond model tuning and focus on one of the most foundational aspects of AI performance: data.

For more information, examples, and usage guidance, visit the official `APPFL Data Readiness documentation <https://appfl.ai/en/latest/tutorials/examples_dr_integration.html>`_.

Citation
--------

This collaboration is described in the following publication:

.. code-block:: bibtex

   @inproceedings{10.1145/3676288.3676296,
      author = {Hiniduma, Kaveen and Li, Z. and Sinha, A. and Madduri, Ravi and Byna, Suren},
      title = {CADRE: Customizable Assurance of Data Readiness in Privacy-Preserving Federated Learning},
      year = {2025},
      publisher = {IEEE},
      url = {https://arxiv.org/pdf/2505.23849},
      booktitle = {Proceedings of the 21st IEEE International Conference on e-Science (e-Science '25)},
   }