Automating Continuous Compliance for Scalable, Sustainable SaaS ML
summary
Oracle Cloud builds end-to-end ML tooling: touchless CI/CD using GitLab/Bamboo and Terraform, deployment to Nvidia GPUs and bare‑metal, and automated compliance gates that generate unit/regression tests and run simulated XSS/DoS checks. IT support is 70% automated by LLM assistants that resolved three recent access tickets. SLAs target 99.99% uptime with a 500 ms notebook‑launch goal, alarms trigging at ~300 ms while expected response is 200–250 ms. ML-driven right-sizing via a cloud guard autoscaler shrinks underutilized instances and reduces carbon and cost.
AI-written summary · about
chapters (12)
- 0:00Welcome to 42 Machine Learning 2026
- 0:36Meet Puja Ranjan: Oracle Cloud & ML Engineering Background
- 2:28What Puja’s Team Builds: End-to-End ML Tools on the Cloud
- 4:01Why AI in Silicon Valley Matters: Automating IT Support with LLMs
- 6:59Cloud Economics & Empowering Small Businesses with SaaS
- 8:46From Data to Features: How AI Helps Define What Customers Want
- 11:28Building the Feature: Architecture, Compliance, and Secure Dev
- 13:49Deploying at Scale: Touchless CI/CD, Rollouts, and Regulations
- 16:03SLA Monitoring & ML-Driven Reliability (Latency, Alarms, Root Cause)
- 19:10AI in CI/CD Gates: Code Coverage, Security Tests, and Compliance Automation
- 21:05Sustainability & Hardware Optimization with ML (Right-Sizing, Carbon Footprint)
- 22:43Wrap-Up: The Promising Future of AI/ML Integration
related talks
24mFrom Model to Production: Automating the ML Lifecycle with CI/CD
17mContinuous deployment for ML: The new software development lifecycle
20mPredictive Analytics for 99.98% Oracle Cloud Availability
26mManaging ML Models @ Scale
12mAI-Powered Predictive Analytics for Multi-Cloud Management
19mReliability at Scale: How AutoML is Transforming Enterprise SRE
26mRunning Secure AI/ML Experiment in a Unified MLOPS-DevSecOps Pipeline
10mAutomation of decisions to augment efficiency in the enterprise