AI as a Service
summary
Defines AIaaS as hosting models in the cloud for pay‑as‑you‑go access and claims it cuts cost and speeds deployment; demonstrates a live case where an e‑commerce chatbot on Microsoft Azure Cognitive Services handles routine queries, raises satisfaction, and shifts humans to complex cases. Shows building a Python API (character counter) and packaging with Docker + Kubernetes, discusses Amazon SageMaker for heavy ML, a Lambda→S3→SageMaker image segmentation pipeline (suggests segmenting every 10 frames), and recommends RAG plus CI/CD, monitoring (Prometheus/Grafana) and model versioning.
AI-written summary · about
chapters (28)
- 0:00AI as a service intro
- 1:49Defining AI as a service
- 6:46Why orgs adopt AI as a service
- 9:01Benefits and challenges
- 11:35AI methods and platforms
- 13:22AI services like vision and NLP
- 15:35AI in e-commerce support
- 20:34Real-world AI applications
- 31:03Conclusion
- 36:07Serving AI via APIs
- 41:54AI model APIs intro
- 42:16Selecting API frameworks
- 43:41Model prep and serving steps
- 46:11API endpoints and testing
- 46:55Monitoring and deploying APIs
- 51:36Advanced AI serving topics
- 55:52Hands-on API serving example
- 1:08:43Break and recap
- 1:12:08Docker for AI deployment
- 1:13:33VMs to containers
- 1:14:55What is Docker
- 1:16:01VMs vs Docker
- 1:17:25Docker key features
- 1:18:50Orchestration and CI/CD
- 1:21:00Hands-on Docker example
- 1:27:58Q&A on Docker and jobs
- 1:31:58AI career paths
- 1:40:42Final Q&A
related talks
15mAI as a Service: The Next thing for Big Tech
27mAI as a Service Scaling AI with Ease and Innovation
43mAI Knows the Board Better Than You Do
31mAzure AI Innovations at Microsoft
30mAI and API, made for each other
29mAI for All
12mAI as a Force Multiplier for Cloud Native Enterprise Development
01h 29mRunning AI Application in Minutes w/ AI Templates