Introduction to Cloud Computing

About

  • Instructor: Yogesh Simmhan (www | mail)
  • TAs: Srinidhi, Roshini, Shreemaye, Manjil, Teams Copilot Agent
  • Course Code: DS 252
  • Credits: 3:1
  • Semester: Aug, 2026
  • Schedule:
    • Lectures: TTh 1000-1130AM
    • First class: Mon 11 Aug, 2026, Laptop mandatory.
    • Tutorials: Fridays, TBD
  • Room: CDS 202
  • Teams Link: DS252: Introduction to Cloud Computing (Aug 2026) [join code: gr2tpt7]
  • Previous edition: [Aug 2025]

Pre-requisites

  • Knowledge of Data Structures, Programming and Algorithm concepts.
  • Programming experience required, preferably in Python, Java or C++.
  • One of the following courses or prior instructor approval is required: DS 221 (Introduction to Scalable Systems), E0 251 (Data Structures and Algorithms), DS 295 (Parallel Programming), E0 225 Design and Analysis of Algorithms, E0 253 (Operating Systems) or equivalent.
  • PG/PhD students who have a BE/BTech degree in Computer Science but none of the above courses may take this course with prior approval of the instructor.
  • UG students who have taken UMC 201 (Data Structures and Algo) but none of the above courses may take this course with prior approval of the instructor.

Introduction

Cloud computing is a key distributed systems paradigm that has grown popular in the last two decades. Cloud technologies are pervasive, touching our daily lives any time we access the Internet, use an AI chatbot, or make a purchase on an app. Most enterprise applications are all cloud-native. While innovative Cloud services are offered by major providers, Cloud computing is also grounded in foundational distributed systems and scalable software systems principles, and is an active area of research by the academic community.

This hands‐on course on Cloud computing will teach both the fundamental concepts of how and why Cloud systems work, as well as Cloud technologies that are based on these concepts. Spread across multiple modules, you’ll gain familiarity with virtualization and distributed architectures, then dive into the service models (IaaS/PaaS/SaaS) that power Cloud vendors such as Amazon AWS, Microsoft Azure and Google GCP. Then you will be introduced to cloud storage services, followed by computing and serverless orchestration for cloud and edge computing. You will then learn about cloud observability, security and manageability. Lastly, you will get to learn specialized topics on MLOps, LLMOps, quantum computing in the cloud and AI‐driven agentic workflows using platforms such as Bedrock, LangGraph and similar cloud-native agent frameworks. These should give you a comprehensive overview of cloud computing to help you understand, design and build cloud applications.

Each week combines an instructor overview and summary/Q&A, AI-driven in-class learning using Teams Copilot and peer-based learning, along with regular hands‐on lab tutorials. You will also take on an ambitious capstone project where teams assisted by AI agents architect, deploy and secure a full-stack cloud-based application.

Students who perform well in this course will be eligible to undertake their final year M.Tech./UG project in the DREAM:Lab under the instructor’s supervision.

AI-based Teaching Methodology

This course uses a new teaching paradigm: AI Agents we have designed using Teams Copilot will serve as your primary teaching instructor for the various modules that have been designed. This improves upon last years’ experience with a similar format.

  1. You will interact directly with the Teams Copilot Agents as intelligent and personalized study partners to learn the topics for the course. This will be both in-class during lecture hours as well as offline at your own pace. Laptops are mandatory in class.
  2. You will occasionally engage in peer learning cohorts among your fellow students in class to better grasp your understanding from the Agent.
  3. The (human) instructor (Yogesh) will offer an overview and summary of each module, help fill your gaps in understanding through Q&As, be available to guide you in your learning, and conduct assessment. (Human) TAs will lead regular lab sessions for hands-on activities for each module. Laptop is required.
  4. You will also have a major capstone programming project as a team, including active use of AI agents to build a complex cloud-based application. You will use AI agents to design your application requirements, write a bulk of the code, and test it.

Students should be open to this non-traditional modality of study. You’ll experiment with novel interactive workflows, share real-time feedback, and shape the evolution of our teaching methods. As early adopters of AI-driven learning, you’ll not only master state-of-the-art cloud concepts and technologies, you’ll also help define how we educate the next generation of students in higher education.

We will also conduct regular anonymous surveys to gauge the feedback from the students for this new learning approach and include that to improve the teaching process.

Learning Modules

  • Module 1: Systems Foundations and Cloud Deployment Models
    • Week 1: Virtualization & Container Runtimes Fundamentals
    • OS abstractions, hypervisors, VMs, Linux namespaces, cgroups, Docker/OCI containers, and performance trade-offs between VMs and containers.
    • Week 2: Cloud Service & Deployment Models
    • IaaS, PaaS, SaaS, public/private/hybrid clouds, elasticity, scalability, shared responsibility, tenant isolation, and cross-cloud provisioning.
  • Module 2: Cloud Computing, Storage and Orchestration
    • Week 3: Cloud Storage, Scaling & Cost Models
    • Block, file and object storage, HDFS/Ceph/MinIO, managed cloud storage services, lifecycle policies, replication, autoscaling, pricing models, and FinOps.
    • Week 4: Serverless & Event-Driven Architectures
    • FaaS and BaaS models, pub/sub, queues, event sourcing, CQRS, workflow orchestration, reliability patterns, cold starts, and serverless cost analysis.
    • Week 5: Kubernetes Orchestration & Service Mesh
    • Kubernetes control plane, pods, deployments, StatefulSets, Helm, CRDs, operators, Envoy/Istio service mesh, and observability using Prometheus and Grafana.
  • Module 3: Cloud Manageability
    • Week 6: DevOps & Infrastructure as Code
    • CI/CD, GitOps, Terraform, reusable infrastructure modules, remote state, GitHub Actions, and automated cloud infrastructure deployment.
    • Week 7: Security, Compliance & Observability
    • IAM, cloud network security, least privilege, compliance frameworks, metrics, logs, traces, SLOs/SLIs/SLAs, error budgets, and chaos engineering.
  • Module 4: Emerging Topics on Cloud
    • Week 8: Cloud-Native MLOps, LLMOps & Agentic AI Platforms
    • MLOps lifecycle, feature stores, Kubeflow/SageMaker pipelines, model deployment, drift detection, retraining, LLMOps, and agentic platforms such as Bedrock, LangGraph and related cloud-native agent frameworks.
    • Week 9: Edge Computing & the Cloud Continuum
    • Edge, fog, MEC and central cloud architectures, K3s, AWS IoT Greengrass, data synchronization, edge AI inference, quantization, federated learning, and device management.
    • Week 10: Quantum Computing in the Cloud
    • Quantum computing concepts, cloud-hosted quantum platforms, hybrid quantum-classical workflows, and emerging applications of quantum services.
    • Week 11: Multi-Cloud & Hybrid Strategies
    • Multi-cloud and hybrid-cloud architectures, workload placement, interoperability, portability, governance, cost trade-offs, and operational challenges across cloud providers.

Assessment

Given the AI-driven learning paradigm, the assessment will also be non-traditional. This is the tentative evaluation approach. This may be revisited during the term based on outcomes.

  • Study summary of AI-driven learning: You will submit the Teams Copilot transcripts of the conversations you have had with our AI instructor during class and outside of class. This will be evaluated for the level of information you were able to elicit from the agent, your engagement in the learning process, and self-assessment quizzes you take. You will also occasionally form peer-groups to collaborate with other students to help understand the module better. You will also submit surveys on for each lecture your learning experience. Your weekly tutorials will also be graded. Up to 4 points per module will be given based on AI-driven learning inside and outside class lectures, with 1 point per module of extra credit. [4 modules * 3 points for lectures + 4*1 point for tutorials = 16 points + up to 4 points extra credit for off-class learning]
  • Quizzes: You will have a quiz at the end of each module to evaluate your understanding of the module. One warm-up quiz will be adaptive based on your learning progress while another is a standardized quiz. [4 modules * (2+6) points = 32 points]. We will also have one final exam [22 points]. These will all be closed-book proctored in-class exams. We will have an instructor-led lecture to revise topics after each quiz based on performance.
  • Capstone Project: Student teams of 2-4 will complete a major cloud application development project with the active assistance of AI agents. Given the access to AI agents, this project will be ambitious in nature. It will evaluate your ability to design, build and test large cloud apps with the active involvement of LLMs. Sample project topics will be provided or you can propose your own. You will be evaluated on the specification/design of the application [5 points], a final demo/presentation/viva [20 points], and red-team testing of other teams’ applications [5 points], for a total of [30 points]. You will have to answer a viva on your understanding of the code and design. An industry panel will review the final demo and presentation.

Attendance

Attendance will be taken for this course. You will need a minimum attendance of 75% for the in-class lectures. Students must bring a laptop to class.

Academic Integrity

Students must uphold IISc’s Academic Integrity guidelines. Failure to follow them will lead to penalties.

As part of the learning process you are required to the the AI agent provided by the course. You should only submit the transcripts of your own Teams Copilot chat with our AI agent, without any changes to them. You may also use additional sources for your learning.

All quizzes and exams must be answered in-class, independently by you, without any assistance by other humans, AI agents, Internet, books, notes, etc.

Project work should be done with the active help of AI agents. You will also submit a summary of work done by each of the team members, and by the AI agents. You may not directly copy code from external sources and websites. Viva is part of project grading.