Expert Insights on Developing Safe, Secure and Trustworthy AI

Charles Vardeman joins us from the University of Notre Dame where he is a Computational Scientist in the Center for Research Computing and a Research Assistant Professor of Computer Science Engineering. He will share with us his team's Trusted AI (TAI) lessons learned two years into the course of their US Federally funded TAI research projects. The team created a framework for Trustworthy AI that automates setup so that developers and decision-makers can focus their efforts on strategy and innovation instead of wrangling the complexities of setting up such a system. This framework covers, Privacy, Sustainability, Accountability, Explainability, Safety and robustness, and Fairness issues at every part of the operational environment. It leverages GitOps and a Data-centric approach as foundational components to automate and streamline the development pipeline and achieve Trustworthy AI. Charles is a member of CI Compass which provides expertise and active support to cyberinfrastructure practitioners at NSF Major Facilities in order to accelerate the data lifecycle and ensure the integrity and effectiveness of the cyberinfrastructure upon which research and discovery depend. Important links from Charles's presentation: Link to Charles's slides: https://drive.google.com/file/d/1_MhjxqKUJFjW16aoMzfN9brCYDmDkvQP/view?usp=sharing Link to KD Nuggets article: https://www.kdnuggets.com/expert-insights-on-developing-safe-secure-and-trustworthy-ai-frameworks Paco Xander Nathan Video: https://www.youtube.com/watch?v=ScIHpAhPccM Paco Xander Nathan Slides: https://derwen.ai/s/mqqm#1 SPDX 3 Model: https://github.com/spdx/spdx-3-model SBoM Tool metadata analysis using Knowledge Graphs Using the TAI Framework: https://nd-crane.github.io/sbom-analysis/ DVC-Fast.ai Integration: https://dvc.org/doc/dvclive/ml-frameworks/fastai Fast.ai: https://docs.fast.ai/ DVC-Hugging Face integration: https://dvc.org/doc/user-guide/integrations/huggingface Dataset Cards: https://huggingface.co/docs/hub/en/datasets-cards Model Cards: https://huggingface.co/docs/hub/en/model-cards Croissant: https://github.com/mlcommons/croissant Croissant Metadata file on Coco 2014 dataset: https://github.com/mlcommons/croissant/blob/main/datasets/1.0/coco2014-mini/metadata.json AI Alliance: https://thealliance.ai To learn more about Iterative's open-source and SaaS tools please visit: 🧑🏽‍💻 *Our free online course:* [https://learn.iterative.ai](https://learn.iterative.ai/) ✍🏼 *Our docs:*https://dvc.org/doc (Data Version Control, Pipelines, Experiments) [https://studio.iterative.ai](https://studio.iterative.ai/) (Team Collaboration, Experiments, Model Registry) *Try out the DVC Extension for VS Code here:* https://marketplace.visualstudio.com/items?itemName=Iterative.dvc *Join the Community on our Discord server:* https://discord.gg/W49xzNmycw #dvc #machinelearning #datascience #generativeai