Machine Learning Operations (MLOps) Specialization Course

Program Curriculum:

Module 1:

MLOps Introduction

• What is MLOps?
• Machine learning industrialisation challenges
• MLOps Motivation: High-level view
• MLOps challenges
• MLOps challenges similar to DevOps
• MLOps Components
• Machine Learning Life Cycle
• How does it relate to DevOps, AIOps, ModelOps, and GitOps?
• Major Phases - what it takes to master MLOps
• CI/CD in Production Case Study

Module 2:

Introduction to ML and MLOps stages

• MLOps Maturity Model
  
• Detailed MLOps and stages
• Versioning
• Testing
• Automation (CI/CD)
• Reproducibility
• Deployment
• Monitoring
• MLOps Architectures

• Architectures - Open Source tools - Kubeflow, MLFlow, Metaflow, Kedro, ZenML, MLRun, CML
• Architectures - Cloud Native tools - AWS, GCP and Azure
• Comparison among cloud-native tools
• The cost-benefit approach of each architecture and MLOps maturity
• List of tools involved in each stage (MLOps tool ecosystem)
• Different Roles involved in MLOps ( ML Engineering + Operations )

Module 3:

Introduction to Git [Hands-on]

• Overview of Git
• Understanding branching strategies and REPO
• Standard GIT branching strategies(development, feature, bug, release, UAT)
• Practising important Git commands
• GitHub Action overview and working

Module 4:

Introduction to CI/CD [Hands-on]

• Introduction to CI and CD
• CI/CD challenges in Machine Learning
• Steps involved in the CI/CD implementation in ML lifecycle and workflow
• A glimpse of popular Tools used in the DevOps ecosystem on the Cloud.

• AWS DevOps
• AWS CodeCommit
• AWS CodePipeline
• AWS CodeBuild
• AWS CodeDeploy
• Project: AWS DevOps Pipeline
• GCP DevOps

• Cloud Source Repositories
• Cloud Run
• Cloud Build
• Cloud Deploy
• Artifacts Registry
• Project: GCP DevOps Pipeline
• Azure DevOps
• Azure Boards
• Azure Repos
• Azure Pipeline
• Azure Test Plans
• Azure Artifacts
• Project: Azure DevOps Pipeline

Module 5:

Docker & Kubernetes Overview [Hands-on]

• Docker Foundation
• Installing docker on Windows, macOS & Linux
• Managing Container with Docker Commands
• How does it work? Docker registry - Docker Hub
• Building your own docker images
• Project: Deploy ML model in docker container
• Kubernetes Overview
• Kubernetes Architecture
• Nodes
• Control Plane
• API Server
• Kubernetes Resources
• Pod
• ConfigMap
• Service
• Secret
• Ingress
• Deployment
• StatefulSet
• DaemonSet
• Volumes (PVC)
• Minikube
• Project: Deploy ML model in Kubernetes cluster

Module 6:

Kubernetes Deployment Strategy [Hands-on]

• Monitoring
  
• Liveness and Readiness Probes
• Labels and Selectors
• Project: Deploying an ML Model using Docker and Amazon EKS

Module 7:

Introduction to Model Management [Hands-on]

• What is a Model Management
• What are the various activities in Model Management
• Data Versioning
  
• Code Versioning
• Experiment Tracker
• Model Registry
• Model Monitoring
• A high-level overview of the below Model Management tools
• MLFlow
• Project: Deploy MLFlow stack on the cloud
• Project: Build, train, and deploy an ML model using MLFlow Experiments and MLFlow model registry.
• DVC
• Git Large File Storage (LFS)

Module 8:

Feature Store [Hands-on]

• Introduction to Feature Stores, SageMaker Feature Store, Vertex AI Feature Store, Databricks, Tecton, Feast, Hopsworks etc.
• Feast open source feature store
• Feature Store: Onlne Vs Offline
• Project: Deploy Feast Online/Offline feature store
• Online Feature Store using DynamoDB
• Offline Feature Store using S3
• Monitor ML features using Amazon SageMaker Feature Store and AWS Glue DataBrew
• Monitor features programmatically
• Visualizing feature drift over time

Module 9:

Cloud ML Services 101 [Hands-on]

• AWS SageMaker
• Introduction to Amazon Sagemaker
• Using Amazon S3 along with Sagemaker
• Amazon SageMaker Notebooks
• Notebook instance type, IAM Role & VPC
• Build, Train & deploy ML Model using Sagemaker
• Endpoint & Endpoint configurations
• Generate inference from deployed model
• AWS SageMaker Pipelines
• SageMaker Studio & SageMaker domain
• SageMaker Projects
• Repositories
• Pipelines & Graphs
• Experiments
• Model groups
• Endpoints
• Project: Deploy an end-to-end MLOps pipeline using Sagemaker Studio.
• GCP VertexAI
• Introduction of Vertex AI
• Gather, Import & label datasets
• Build, Train & deploy ML Solutions
• Manage your models with confidence
• Using Pipelines throughout your ML workflow
• Adapting to changes in data
• Creating models with Vertex AI and deploying ML models using AI platform pipelines
• Project: Deploy an end-to-end MLOps pipeline using Vertex AI
• Azure MLOps
• Azure Machine learning studio
• Azure MLOps
• Azure ML components
• Azure MLOps + DevOps
• Fully automated end-to-end CI/CD ML pipelines
• Project: Deploy an end-to-end MLOps V2 pipeline using Azure Machine Learning

Module 10:

Kubeflow Intro [Hands-on]

• Kubeflow Introduction
  
• Kubeflow- Who uses it
• Kubeflow features
• Kubeflow Fairing
• Kubeflow Pipelines
• Kubeflow use cases
• Project: Pipeline formation with Kubeflow
  

Module 11:

Introduction to Model Monitoring [Hands-on]

• Importance Of Model Monitoring
  
• What are the various types of monitoring related to the model
• The architecture of monitoring ecosystem in AWS/Azure/GCP
• AWS Model Monitoring
• Azure Model Monitoring
• GCP Model Monitoring
• Optimize and Manage Models at the Edge
• Common Issues in ML Model Deployment
• Feedback Loop Role
• Project: Model & infrastructure monitoring using cloud tools

Module 12:

Introduction to Automl tools [Demo]

• H20 MLOps
  
• Valohai
• Domino Data Lab
• neptune.ai
• iguazio
• W&B

Module 13:

Post-Deployment Challenges [Hands-on]

• Post Deployment Challenges intro
  
• Post Deployment Challenges - ML Related
• Challenges when deploying machine learning to edge devices
• Post Deployment - Monitoring the Drift - Evidently
• Monitoring the Drift - Using Sagemaker
• Post Deployment Challenges - Software Engineering Related
• Common Issues in ML Model Deployment
• Project: Evidently AI for Monitoring the Drift

About Program

Looking for getting started with Hands-on Machine Learning Operations (MLOps) with a real-time

the project, then you've come to the right place. As per the market survey, 2023 is the year of MLOps and

would become the mandate skill set for Enterprise ML projects.

Corporates have been experimenting with machine learning models for a long time, 85% of Machine

Learning projects do not reach production. In addition, the MLOps have exponentially grown in the last

few years. MLOps was estimated at $23.2 billion for 2019 and is projected to reach $126 billion by

2025. Therefore, MLOps knowledge will give you numerous professional opportunities helps

organizations to bring in real business value

About Psitron Technologies:

Psitron Technologies is an IoT and AI company. Our mission at Psitron is to connect the world with innovative technologies. Psitron has responsibility for developing innovative innovations for addressing current problems in industries, especially focusing on industry 4.0 solutions.

Key Highlights

• 60 Hours of Live sessions from Industrial Experts
• 50+ Live Hands-on Labs
• 5+ Real-time industrial projects
• One-on-One with Industry Mentors

Who Can Apply for the Course?

• Data Scientists
• Data engineers & Data Analysts
• Research/Applied Scientists
• ML engineers
• DevOps engineers
• Aspiring MLOps Professionals and Enthusiasts
• Machine Learning professionals who want to deploy models to production
• Anyone who wants to learn Docker & Kubernetes, AWS, Azure, GCP, DVC, Feast, MLFlow etc
• Individuals interested in the data and AI industry

Hottest Job of the 21st Century

Machine learning and artificial intelligence is a field that drives major innovations across different

industries. It is predicted that in 2023, the AI market will reach $500 billion, and in 2030, $1,597.1 billion

in size. This means that machine-learning technologies will continue to be in high demand in the near

future.

Schedule:

• Starting from 15th April, every weekend
• One-on-one debugging 

Note: All sessions are on Weekends

• 7:00 PM to 11 PM IST
• 8:30 AM to 12:30 PM EST
• 2:30 PM to 6:30 PM CET

Group Discount

Register as a group of 2 or more and get 10% OFF per person on registrations.

Certification

Participants will be certified by Psitron Technologies Pvt. Ltd.

Contact Us:

Sarathkumar. C

Founder & CEO

Psitron Technologies Pvt. Ltd.

+918940876397 / +918778033930

sarath@psitrontech.com

www.psitrontech.com

P.S. - Today, don't play it safe, play it smart.