Course Overview
- Course Title: Certified Infra AI Expert: End-to-End GPU-Accelerated AI
- Instructor: Vivian Aranha (AI Specialist, Principal Engineer, IT Professional Trainer)
- Target Audience:
- AI engineers, ML developers, and system architects
- Professionals working with NVIDIA GPUs (A100, H100, L4, Jetson)
- Cloud/edge AI solution designers
- DevOps engineers integrating AI pipelines
- Enterprise IT teams deploying GPU-accelerated AI
- Prerequisites:
- Basic AI/ML concepts (training, inference, deployment)
- Familiarity with Linux command-line (Ubuntu recommended)
- Basic Docker/containerization knowledge (covered in-course if lacking)
- Access to NVIDIA GPU (A100/H100/L4/Jetson) or cloud GPU instance (AWS/Azure/DGX Cloud)
- Stable internet for NGC containers, SDKs, and pretrained models
Curriculum Highlights
-
Key Topics Covered:
- NVIDIA AI Enterprise Stack: Full workflow from setup to deployment
- GPU-Accelerated Infrastructure:
- NVIDIA drivers, Kubernetes GPU nodes, Helm charts
- AWS/Azure/DGX Cloud integration
- Model Optimization & Deployment:
- TensorRT for high-performance inference
- TAO Toolkit (transfer learning, quantization)
- Triton Inference Server (scalable deployment)
- Real-Time AI Applications:
- DeepStream SDK (video analytics, sensor fusion)
- RAPIDS (GPU-accelerated data processing)
- Cloud-Edge & Digital Twin Integration:
- Omniverse for simulation
- IoT sensor fusion with Jetson
- Security & Compliance:
- Container security, NVIDIA License Server, enterprise licensing
- Vertical SDKs:
- Metropolis (smart cities)
- Riva (speech AI)
- NeMo (NLP)
- Clara (healthcare AI)
- Merlin (recommender systems)
- Capstone Projects:
- Video surveillance (DeepStream)
- Digital twin simulation (Omniverse)
- Smart edge AI (Jetson + IoT)
-
Key Skills Learned:
- Architect end-to-end GPU-accelerated AI pipelines
- Optimize models with TensorRT and quantization
- Deploy real-time AI using DeepStream, RAPIDS, Triton
- Integrate AI with cloud (Kubernetes, Helm), edge (Jetson), and digital twins (Omniverse)
- Apply enterprise security, licensing, and DevOps CI/CD for AI
- Earn Certified NVIDIA AI Expert credential
Course Format
- Duration:
- 2.5 hours on-demand video
- 11 articles
- 11 downloadable resources (labs, SDKs, configuration guides)
- Format:
- Self-paced online course (lifetime access)
- Hands-on labs with real-world projects
- Mobile & TV access
- Resources:
- Downloadable NGC containers and pretrained models
- Configuration templates for Kubernetes, Helm, Triton
- Code repositories for DeepStream, TensorRT, TAO
- Certificate of completion
Additional Information
- Certification: Certified NVIDIA AI Expert (industry-recognized credential)
- Industry Applications:
- Autonomous systems, healthcare AI, financial modeling
- Smart cities, manufacturing, IoT analytics
- Digital twins, cloud-edge hybrid AI
- Instructor Stats:
- 4.2/5 rating (532 reviews)
- 85,109 students enrolled across 6 courses