Creating a fair and trusted AI enterprise

on YouTube ↗12m

" summary

Enterprise AI starts with a concrete use case and a clean training dataset, not headlines about deep nets. Deployment and data prep take more time than model development; discovery, profiling, provenance and governance are the critical work. IBM case work routes incidents for a 3,000‑developer org and uses features (days open, code component, customer history) to prioritize fixes, yielding a 40% NPS improvement and selective regression runs. Tooling includes Watson OpenScale and governance for model fairness and explainability.

AI-written summary · about

" chapters (8)

1. 0:00Fair and trusted AI intro
2. 0:22Clear AI use case
3. 1:45AI: data, development, deployment
4. 3:11IBM AI efficiency apps
5. 6:42Trust and fairness imperatives
6. 8:07Open-source AI ethics
7. 8:49AI needs information architecture
8. 10:16Enterprise AI challenges

" related talks

• 27mAI at Scale in the EnterpriseClarisse Taaffe Hedglin · OpenPOWER Summit NA 2020
• 12mTrust and transparency of AI for the enterpriseRuchir Puri · Artificial Intelligence Conference in London 2018
• 30mEnterprise AI: How To Operationalize Data Analytics At ScaleRoman Swoszowski · Big Data LDN 2022
• 33mResponsible AI in Practice: Real-World Examples and ChallengesRay Eitel-Porter, Mina Saidze, Björn Bringmann, Steffen Bosse · WeAreDevelopers World Congress 2024
• 56mAI War Stories: Real world learnings delivering enterprise AI solutionsAdam Stephenson · NDC Melbourne 2025
• 37mBuilding a case for Fair and bias-free AIChan Naseeb · International Data Science Conference 2021
• 39mBig Data Information Architecture for AIToby Woolfe · Big Data Conference Europe 2019
• 09mUnlocking the value of your dataDaniel Hernandez · Strata Data Conference 2019 - New York, NY