Realizing value and business outcomes with AI
summary
Ferrari and partners show concrete steps: using SageMaker and Bedrock plus Stable Diffusion to generate rare crash images, raising a computer‑vision pipeline from ~72% manual accuracy to ~84% automated and then another ~4–5 percentage points after synthetic augmentation; claim that 70–80% of models never reach production and a Fortune 200 survey finds 91% of impediments are people/process/culture. They present a hardened platform: collaborative workbench, ISO‑style ISO-26262 verification/versioning, multi‑tenancy and moving compute to data, and say smaller edge models plus governance drive industrialization.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 0:20Welcome
- 0:43AI driven outcomes
- 2:34How will business perceive AI
- 6:20What are the potential of machines
- 6:48Barriers to AI adoption
- 9:42democratizing value creation
- 11:10generative AI use cases
- 15:44data science hub
- 18:15dont force fit
- 18:59Ferrari
- 22:07Two more trends
- 23:42AWS AI strategy
- 26:47Autonomous driving
- 31:02Future of AI
- 39:27Watch outs
related talks
01h 00mSecurely accelerating generative AI innovation
30mGenerative AI in action: from POC to business value
23mDelivering business success with AI
38mBuilding AI responsibly with a GRC strategy
27mPractical AI: Driving Tangible Value in the Enterprise
36mSolve common business problems with AI/ML
56mPanel: Scaling AI successfully
27mAI at Scale in the Enterprise