Story of Evolution of MLOps, LLMOps and AgenticAIOps

Getting started with the Use Case and Environment Setup

0/12

Use Case – House Price Predictor – Regression

03:20

07:06

Understanding End to End ML Practices and MLOps

19:30

Environment Setup Overview

09:24

Setting up Docker / Podman with Compose

04:40

Launching MLflow for Experiemnt Tracking

07:31

Understanding the Project Directory and Scaffold

07:46

Setting up Python Virtual Environment with UV

06:08

Working with Jupyter Notebooks

06:54

Lab Guide – Environment Setup

03:48

Getting Started with Use Case and Environment Setup

From Raw Data to Models

Understanding Data Science with Feature Engineering and Experimentation

0/12

Learning Data Engineering

03:22

12:54

Experimental Data Analysis

08:08

Understaing Feature Engineering Concepts

05:51

Building New Features for House Price Predictor

05:25

Preparing for Model Experimentation

06:18

Data Splitting with x_train, y_train, x_test, y_test

04:58

Defining Algorithms and Hyperparameter Grids

05:34

Running Model Experiments to find the Best Model and Hyperparamters

From Raw Data to Models

Packaging Model along with FastAPI Wrapper and Streamlit with Containers

0/12

Handover from Data Scientist to ML Engineer / MLOps

03:19

06:36

Running Feature Engineering and Preprocessing Jobs

04:41

Building and Training Final Model with Configs from Data Scientists

06:04

Wrapping the Model with FastAPI with Streamlit Client Apps

06:47

Writing Dockerfile to package Model with FastAPI Wrapper

17:47

Debugging and Fixing Image Failures, Launch and Validate FastAPI

11:21

Packaging and testing Streamlit App

09:30

Packaging and Model Serving Infra with Docker Compose

14:53

Packaging Model with FastAPI, Streamlit, and Containers

04:21

DAGs, GitHub Actions and our MLOps CI Workflow

Lab Guide

Setting up MLOps CI Workflow with GitHub Actions

0/12

Moule Intro

02:58

12:40

Understanding GitHub Actions Syntax

07:39

Writing an executung out first GitHub Actions Workflow

13:50

Adding Data and Feature Engineering Steps with Model Training

05:53

Model Training Step with MLFlow for Tracking

07:43

Adding Image Build and Publish Step with Docker

05:53

Configurating Registry Token and publishing Image to DockerHub

06:50

Modular, Multi Stage MLOps CI Workflow Pipeline

Setting up MLOps CI Workflow with GitHub Actions

Building Scalable Prod Inference Infrastructure with Kubernetes

0/13

Designing Scalable Infrastructure for Model Inference

02:00

04:35

Introduction to Kubernetes for Machine Learning

10:28

Kubernetes Core Concepts – Pods, Deployments and Services

09:13

Simplest way to build a 3 Node Kubernetes Cluster with KIND

10:04

Deploying Streamlit Frontent App with Kubernetes

12:04

Exposing the Streamlit App with Kubernetes NodePort Service

06:24

Creating Deployment & Service for the Model wrapped in FastAPI

07:10

Connecting Streamlit with Model using Kubernetes native DNS Based Service Discovery

11:09

Easy way to Generate Kubernetes Manifets and YAML

06:30

Scalable Inference Infrastructure with Kubernetes

01:58

Installing Prometheus and Grafana with Helm

Lab Guide

Upcoming : Inference Serving with Kubernetes (KubeRay/Kubeflow/BentoML)

Monitoring and Autoscaling a ML Model – Prometheus, Grafana, KEDA, HPA, VPA

0/16

06:05

Exploring Monitoring Metrics with Grafana and Prometheus

08:16

Adding Instrumentation for FastAPI along with Custom Dashboard

11:29

Automatic Capacity Scaling Concepts

04:01

Installing KEDA and Configuring Resource Spec

05:22

Configuring Scaled Objects with KEDA

07:26

Getting started with Load Testing Model Inference

07:28

AI Based Troubleshooting Monitoring with ChatGPT

08:52

Running Load Test and Analysing Autoscaling

09:36

CPU Based Auto Scaling with KEDA

10:44

Adding a Verticle Pod Autoscaler (VPA)

11:42

Lab Guide – Setting up Monitoring with Prometheus and Grafana

Lab Guide – Autoscaling with KEDA, HPA and VPA