৬ মাসের স্টাডি প্ল্যান

৪৪ টি লাইভ ক্লাস

১০টি রিয়েল লাইফ প্রজেক্ট + ক্যাপস্টোন প্রজেক্ট

প্রোগ্রেস ট্র্যাকিং

উইকলি ৩ দিন সাপোর্টঁ ক্লাস

কমিউনিটি সাপোর্ট

লাইফটাইম এক্সেস

জব মার্কেট গাইডলাইন

ইন্টারভিউ হ্যান্ডবুক

সার্টিফিকেট

ল্যাপটপ/ডেস্কটপ (মিনিমাম ৮ জিবি র‍্যাম)

মিনিমাম Core i3 প্রসেসর

২৫৬ জিবি এসএসডি

ভার্সন কন্ট্রোল সিস্টেম: Git এর বেসিক ধারণা ও ব্যবহার জানা।

লেগে থাকার মানসিকতা

গ্রুপ ডিসকাউন্ট চাই

রিকোয়েস্ট পাঠান

নিচের তথ্য দিন — আমাদের টিম কল করে ছাড়সহ এনরোলমেন্ট বুঝিয়ে দেবে।

বন্ধুদের সাথে এনরোল করুন

একা শিখলে মাঝপথে থেমে যাওয়া সহজ। একসাথে শিখলে কোর্স শেষ করার চান্স অনেক বেশি — আর সবাই বাড়তি ছাড় পায়।

রিকোয়েস্ট পেয়ে গেছি!

বন্ধু বা কলিগদের সাথে একসাথে এনরোল করুন

Docker

Kubernetes

Prometheus

Grafana

Terraform

Git & Git Hub

Linux

Md Shakil Haque

Md. Abu Naeem (ব্যাচ ১১)

Mehedi Hasan Masum (ব্যাচ ১৩)

Mridul Mahmud (ব্যাচ ১৩)

Sina (ব্যাচ ১৩)

Hridoy Sarker (ব্যাচ ১২)

Ornab Barua (ব্যাচ ১২)

Ibnul Quayum Imran

Md. Tanvir Arif Siddiqui

Farhan Sadiq Mullick

Ariur Rahman

MD. Tahsin Ferdous

Md Ebrahim Mia

Kazi Fabliha Tasnim

Shuayeb Hasan

Shibu Chandra Debnath

MD Sarowar Alam

Git Branching, Repository Roles & Secure History Management

Kubernetes Architecture & Cluster Setup

Ashik Mostofa Tonmoy

Security Scanning in CI:CD, Container Vulnerability Assessment, Basic Secrets Management

GitHub Actions দিয়ে CI CD Automation: DevOps MasterClass for Software Developers

Muzahidul Hassan

Mokchhedul Islam

Emon Shil

Habib Hussain

RADWANUL HABIB RATUL

RONY DAS

Badhon Parvej

Abdul Muhaimin Adeeb

SADMAN SANID TANIM

Md. Yousuf Ali

Ali Ahasan

Tamim Abdullah

Abdul Kayum Muslim

Showrav Acharjee

Mohammad Asif Mostafa

Md. Omar Shohid Khan

Md Eashir Arafat

Md. Sihab Uddin

Md Maruf Islam

Md. Sadman Sakib Anik

Md. Ruhul Amin

Kallol Dhar

Syed Nakibul Islam

Ringko K Rahul

MD.MOMINUR RAHMAN

Tutul Chakma

Md. Jakir Hossain

Farhad Hossain

Sumon K. Mondol

Abu Naser Mohammad Moinuddin

ARIFUR RAHMAN ZERIN

Md. Ashraful Islam

Arnab Das

Sidratul Islam

Raisul Islam

Md Zahid Hasan

ব্যাচে ভর্তি হোন

AI Native DevOps: ফান্ডামেন্টাল থেকে অ্যাডভান্সড প্র্যাকটিস সফটওয়্যার ডেভেলপমেন্টের দুনিয়ায় এখন শুধু কোড লিখলেই হয় না। সেই কোড কত দ্রুত এবং নিরাপদে ইউজারের কাছে পৌঁছানো যায়, সেটাই বড় চ্যালেঞ্জ। আর এখন AI নিজেই পাইপলাইন YAML লিখছে, Terraform জেনারেট করছে, ইনসিডেন্ট ট্রায়াজ করছে। কিন্তু আর্কিটেকচারের সিদ্ধান্ত, ট্রেড-অফ আর দায়িত্ব এখনো একজন ইঞ্জিনিয়ারের। এই কোর্সে আপনি টুলস শেখার পাশাপাশি শিখবেন AI-কে সাথে নিয়ে কাজ করা এবং তার প্রতিটি আউটপুট যাচাই করা। পুরোটা হবে একটা রিয়েল Three-Tier অ্যাপ্লিকেশনের ওপর, যেটা প্রথম মডিউল থেকে ক্যাপস্টোন পর্যন্ত আপনার সাথে বড় হবে। এই লাইভ ব্যাচে যা থাকছে:<ul><li>Foundations: শুরু হবে AI ওয়ার্কস্পেস সেটআপ, Linux, AWS EC2, IAM, Nginx, Route53, SSL এবং টিম-লেভেলের Git ওয়ার্কফ্লো দিয়ে। থাকছে ইচ্ছে করে ভাঙা সার্ভার AI দিয়ে ডিবাগ করার Break-Fix ল্যাব। এরপর Three-Tier অ্যাপ্লিকেশন প্রোডাকশনে ডিপ্লয় করবেন।</li><li>Automate and Observe: GitHub Actions দিয়ে CI/CD, যেখানে OIDC দিয়ে কী-লেস AWS অথেনটিকেশন থাকবে। Prometheus, Grafana, Loki ও OpenTelemetry দিয়ে মেট্রিক্স, লগস ও ট্রেসেস দেখবেন। এরপর AWS VPC দিয়ে প্রাইভেট নেটওয়ার্ক, Terraform দিয়ে পুরো ইনফ্রাস্ট্রাকচার কোড হিসেবে লিখবেন এবং Auto Scaling ও AWS CodePipeline দিয়ে High Availability নিশ্চিত করবেন।</li><li>Containers, Kubernetes and Platform: Docker দিয়ে কন্টেইনারাইজেশন ও Trivy দিয়ে ইমেজ স্ক্যানিং। এরপর অ্যাপ্লিকেশন Kubernetes (EKS)-এ মাইগ্রেট করবেন। Helm ও Argo CD দিয়ে GitOps শিখবেন, যেখানে প্রতিটি ডিপ্লয়মেন্ট হবে শুধু Git কমিটের মাধ্যমে। সাথে থাকছে Backstage দিয়ে Internal Developer Platform তৈরি।</li><li>Secure, Operate and AI: DevSecOps ও সাপ্লাই চেইন সিকিউরিটির জন্য SonarQube, Gitleaks, Cosign ও Kyverno। SRE অংশে থাকছে SLO, k6 লোড টেস্টিং ও AIOps। এরপর Serverless ও FinOps। শেষে Agentic DevOps, যেখানে AI এজেন্ট গার্ডরেইল ও হিউম্যান রিভিউসহ রিয়েল ইনফ্রাস্ট্রাকচার পরিবর্তন করবে, আর LLMOps, যেখানে AI ফিচার প্রোডাকশনে চালাবেন।</li><li>AWS Deep Dive: Organizations, SCP, ডিজাস্টার রিকভারি এবং Well-Architected রিভিউ। সাথে AWS Solutions Architect Associate পরীক্ষার প্রস্তুতি।</li><li>Capstone ও ক্যারিয়ার: টিমে বসে পুরো ইনফ্রাস্ট্রাকচার, ডেলিভারি পাইপলাইন আর অবজার্ভেবিলিটি একসাথে তৈরি করবেন। Demo Day-তে লাইভ ফেইলিওর ইনজেকশন ও রিকভারি দেখাবেন। এরপর পোর্টফোলিও, LinkedIn কেস স্টাডি ও মক ইন্টারভিউ দিয়ে ২০২৬-এর জব মার্কেটের জন্য প্রস্তুতি।</li></ul> থিওরি, প্র্যাকটিক্যাল আর AI-এর দারুণ এই কম্বিনেশনে আপনার DevOps ক্যারিয়ার শুরু হোক আমাদের সাথে!

ডকার মাস্টারক্লাস ফর  সফটওয়ার ডেভেলপারস

সাপোর্ট ক্লাস

২০২৭ এ কীভাবে হবেন একজন AI Native DevOps ইঞ্জিনিয়ার?

My First Live Website

Fix the Broken Server

3 Launch and Auto-Deploy a Full App

See Everything: App Health Dashboard

Build the Cloud with Code

Move the App to Kubernetes

Deploy by Git Commit

Self-Service Developer Portal

Secure, Reliable and Cost-Smart Production

Add AI to the App, Safely

Capstone: Production-Ready AI App, End to End

Live Class 01: Course Overview, Lab and AI Workspace SetupTopics: Course roadmap and the single three-tier project thread used across all modules | Cloud account and Git repository onboarding | Setting up an AI coding assistant in the IDE and an AI agent in the terminal | Safe AI usage rules: never paste secrets or credentials into prompts, verify every command, review every change before applying | Running the DevOps demo application locally and exposing it with ngrok Live Class 02: DevOps in 2026: Roles, Skills and How AI Changes the JobTopics: DevOps engineering, SDLC and DevOps culture | 2026 role map: DevOps Engineer, Platform Engineer, SRE, DevSecOps Engineer, MLOps and AI Infrastructure Engineer | What AI now automates (pipeline YAML, boilerplate IaC, first-pass incident triage) and what it does not (architecture judgment, trade-offs, accountability) | Prerequisites: SWE experience, networking knowledge | Soft skills, troubleshooting mindset and patience Goal: Run the project locally, expose it over ngrok, and commit a personal AI usage rules file to the repository. Porject: My First Live Website Assignment: Set up the AI workspace and write a one-page reflection mapping your target role from the 2026 role map.

DevOps in the AI Era: Setup and Role Map

Live Class 01: AWS EC2 and Linux FundamentalsTopics: Introduction to AWS EC2, launching an instance, security groups and key pairs, SSH | Basic Linux commands, package manager (apt), process manager (systemctl) | Linux administration, file permissions and ownership, environment variables | Logs and monitoring, firewall with UFW, cronjobs and task scheduling, troubleshooting basics | AI-assisted log reading: ask the assistant, verify against the actual log, then act Live Class 02: IAM, AWS CLI and Web Server SetupTopics: IAM users, groups, roles and policies, IAM best practices | AWS CLI setup and configuration | Nginx installation, static site hosting, reverse proxy | Route53 and DNS, domain mapping, SSL with Certbot (Let's Encrypt) | AWS Budgets and billing alarms configured on day one Goal: End-to-end production deployment: EC2, UFW, Nginx, Route53, HTTPS. Project: My First Live Website Assignment: Mini project: deploy a static site over HTTPS on your own domain with a billing alarm in place.

Linux and AWS Fundamentals

Live Class 01: Git for Teams: Branching, Reviews and ProtectionTopics: Branching strategies (trunk-based and GitFlow), repository roles | Branch protection, restricting branch and history deletion | Forking workflow, upstream remote setup, pull requests | Conventional commits and pull request templates | AI-assisted pull request review: what to trust and what to check manually Live Class 02: Break-Fix Lab: Debugging Linux and Network Issues with AITopics: Structured troubleshooting method: observe, form a hypothesis, test, confirm | Using an AI agent to diagnose deliberately broken servers (Nginx misconfiguration, DNS, file permissions, firewall rules) | Verifying AI suggestions against logs, man pages and official documentation | Writing a short incident note for each fix Goal: Resolve three deliberately broken server scenarios and document which AI suggestions were correct and which were wrong. Project: Fix the Broken Server Assignment: Submit the three incident notes plus a protected repository with a working pull request workflow.

Git Workflows and AI-Assisted Troubleshooting

Live Class 01: Three-Tier Architecture and Application SetupTopics: Technology stack overview, database deep dive, backend architecture | Implementing the database and backend with required tools Live Class 02: Frontend Development and AWS DeploymentTopics: Frontend architecture, AWS deployment walkthrough | Updating the application and the deployment, full architecture view | Generate an architecture diagram and README with AI, then correct its mistakes Goal: Deploy and manage a production-ready three-tier application on AWS. This application is the project thread for the rest of the course. Project: Launch and Auto-Deploy a Full App Assignment: Three-tier deployment with a reviewed architecture README.

Deploying a Three-Tier Application on a Single Instance

মডিউল ১ থেকে মডিউল ৪ পর্যন্ত যা যা আলোচনা হয়েছে, সেটার উপর ভিত্তি করে থাকবে একটা বড়সড় এসাইনমেন্ট ও লাইভ টেস্ট। ঝালিয়ে নিতে পারবেন নিজেকে।

Exam Week - 1

Foundations (Module 1-5)

Live Class 01: GitHub Actions FundamentalsTopics: CI/CD concepts and benefits, Actions architecture (workflows, jobs, steps, runners) | First pipeline, YAML syntax basics, debugging pipeline failures | Generating and debugging workflows with an AI assistant | Reusable workflows and composite actions | Security review of AI-generated pipelines: pinned action versions, least-privilege permissions Live Class 02: Automate Full Three-Tier DeploymentTopics: Automating deployment of the three-tier application, SSH deploy to EC2 | OIDC federation from GitHub to AWS, removing long-lived access keys | Configuring secrets securely, zero-downtime deployments | Monitoring deployment logs and handling rollbacks Goal: A secure, automated CI/CD pipeline for the three-tier application with keyless AWS authentication. Project: Launch and Auto-Deploy a Full App Assignment: Pipeline with OIDC, pinned actions and a documented rollback.

CI/CD with GitHub Actions for the Three-Tier Application

Live Class 01: Monitoring Fundamentals, Prometheus and OpenTelemetry Topics: Why monitoring matters, Prometheus setup on EC2, Node Exporter | Configuring Prometheus scraping, PromQL | OpenTelemetry concepts: metrics, logs, traces and the OTel Collector Live Class 02: Grafana, Loki, Tracing and Alerting Topics: Grafana setup, Prometheus data source, dashboards, community dashboards | Three-tier application dashboard (backend API, frontend, database metrics) | Centralized logs with Loki | Instrumenting the backend with OpenTelemetry and viewing traces in Grafana Tempo | Writing PromQL and LogQL with AI assistance, then validating queries | Alerts for application health Goal: Full metrics, logs and traces for the three-tier application, with actionable alerts. Project: See Everything: App Health Dashboard Assignment: One dashboard that correlates a slow API request with its trace and logs.

Observability: Metrics, Logs and Traces

Live Class 01: Basic Three-Tier Architecture (Public and Private Subnets)Topics: AWS VPC architecture, custom VPC for the three-tier application | Public and private subnet design, internet gateway, route tables | Frontend in public subnet, backend and RDS in private subnets | NAT gateway, security groups per tier Live Class 02: Production-Ready Fully Private Architecture with Load BalancerTopics: Moving the frontend to a private subnet, AWS load balancers | ALB setup, target groups, exposing the application via ALB/NLB | End-to-end connectivity testing, network security best practices | NAT gateway and data transfer cost awareness Goal: A production-ready, private-network three-tier application behind a load balancer. Project: Build the Cloud with Code Assignment: Network diagram plus a working fully private deployment.

Securing the 3-Tier Application with AWS VPC and Networking

Live Class 01: Terraform FundamentalsTopics: Introduction to IaC, Terraform setup, architecture and workflow | First EC2 instance with Terraform, modules and security | Remote state management (S3) and state locking | Generating Terraform with AI and reviewing it properly: reading the plan, spotting destructive changes, blast radius | Static checks with tflint and Checkov Live Class 02: VPC Architecture with TerraformTopics: Terraform modules for multiple resources, VPC, subnets and networking as code | Load balancer, target groups and security groups as code, bastion host | Managing state and outputs | Terraform plan posted as a pull request comment from CI Goal: Version-controlled, reviewed and policy-checked infrastructure for the full three-tier stack. Project: Build the Cloud with Code Assignment: Terraform repository with CI plan comments and zero high-severity Checkov findings.

Infrastructure as Code: Terraform for the 3-Tier Application

Live Class 01: Auto Scaling for Three-Tier Applications (High Availability and Self-Healing) Topics: AMI creation, launch templates, auto scaling groups | Health checks and self-healing, target tracking and step scaling | Load balancer integration, auto scaling managed with Terraform | Cost impact of scaling policies Live Class 02: CI/CD with AWS CodePipeline to Auto Scaling Groups Topics: CI/CD challenges with auto scaling groups | CodePipeline for backend and frontend auto scaling groups | Blue/green deployment strategy, zero-downtime techniques | Automating pipeline creation with Terraform Goal: A highly available, auto-scaling three-tier system on AWS with CI/CD and IaC. Project: Build the Cloud with Code Assignment: Demonstrate an instance failure being healed automatically with no user-facing downtime.

High Availability Auto Scaling and AWS CodePipeline

মডিউল ৬ থেকে মডিউল ১০ পর্যন্ত যা যা আলোচনা হয়েছে, সেটার উপর ভিত্তি করে থাকবে একটা বড়সড় এসাইনমেন্ট ও লাইভ টেস্ট। ঝালিয়ে নিতে পারবেন নিজেকে।

Exam Week - 2

Automate and Observe (Module 6-11)

Live Class 01: Docker Fundamentals and Containerizing the ApplicationTopics: Containers vs virtual machines, Docker architecture, images vs containers | Dockerfiles for backend and frontend, multi-stage builds, best practices | Pushing images to DockerHub and AWS ECR | Image vulnerability scanning with Trivy | AI review of Dockerfiles: image size, root user, leaked secrets Live Class 02: Docker Compose and Migrating Deployment from VMs to ContainersTopics: docker-compose.yml for the three-tier application, networks, volumes, environment variables | Health checks and container monitoring, VM vs container comparison | Migrating the GitHub Actions pipeline from VMs to Docker, deploying containers on EC2 | Centralized container logging with Loki, monitoring with Prometheus and Grafana Goal: Containerized three-tier application with scanned images, CI/CD, logging and monitoring. Project: Move the App to Kubernetes Assignment: Image under a target size with no critical vulnerabilities.

Dockerizing the Three-Tier Application

Live Class 01: Kubernetes Architecture and Cluster Setup Topics: Container orchestration, control plane and worker components | Cluster setup (Minikube, K3s, EKS), kubectl | Pods, ReplicaSets, Deployments, Services (ClusterIP, NodePort, LoadBalancer), namespaces Live Class 02: ConfigMaps, Secrets and Simple Application Deployment Topics: ConfigMaps, Secrets, environment variables | Deploying and exposing Nginx, rolling updates and rollbacks | Generating manifests with AI, then validating them with dry-run and kubeconform | Inspecting pods and logs, troubleshooting common issues Goal: Kubernetes fundamentals by deploying and exposing an application. Project: Move the App to Kubernetes Assignment: Deploy and roll back an application with a deliberate bad release.

Kubernetes Fundamentals

Live Class 01: Database and Backend Migration to Kubernetes Topics: Migration strategy from Compose to Kubernetes | Provisioning an EKS cluster with Terraform | PostgreSQL on Kubernetes with StatefulSets, PV and PVC | Backend deployment, services, service discovery, configuration and secrets | Database migration and initialization strategies Live Class 02: Frontend Migration, Ingress and Production-Ready Deployment Topics: Frontend deployment and service, Nginx Ingress controller and rules | Horizontal Pod Autoscaler, resource requests and limits | Liveness and readiness probes | Logs with Loki, monitoring with Prometheus and Grafana on Kubernetes Goal: The three-tier application running on EKS with production-grade configuration. Project: Move the App to Kubernetes Assignment: EKS deployment with HPA demonstrated under load.

Migrating the Three-Tier Application to Kubernetes (EKS)

Live Class 01: Packaging the Application with Helm Topics: Why templating: the problem with copy-pasted manifests | Helm chart structure: Chart.yaml, values, templates | Converting the three-tier manifests into a Helm chart | Environment overrides for dev, staging and production | Chart linting and versioning, AI-assisted chart authoring with review Live Class 02: GitOps with Argo CD and Progressive Delivery Topics: GitOps principles: Git as the single source of truth | Installing Argo CD, the App of Apps pattern | Flow: CI builds the image, a Git commit updates the tag, Argo CD syncs the cluster | Drift detection and self-healing | Canary and blue/green with Argo Rollouts | Rollback through Git revert Goal: Every deployment of the three-tier application happens only through a Git commit. Project: Deploy by Git Commit Assignment: Run a canary release that automatically rolls back on a failing metric.

Packaging and Delivery: Helm and GitOps with Argo CD

Live Class 01: Platform Engineering Concepts and Golden Paths Topics: From ticket-based ops to developer self-service | Platform as a product: developers are the customer | Golden paths and paved roads | Backstage: software catalog and TechDocs | Scorecards for service maturity Live Class 02: Self-Service Templates with Guardrails Topics: Backstage software template that creates a repository, CI workflow, Helm chart and Argo CD application | Guardrails built into the template: security defaults, cost tags, ownership | Measuring platform success with DORA metrics | Crossplane overview: self-service infrastructure (concept only) Goal: A developer can create and deploy a new service through the platform without opening a ticket. Project: Self-Service Developer Portal Assignment: Publish a working golden path template and onboard one new service through it.

Platform Engineering:Building an Internal Developer Platform

মডিউল ১২ থেকে মডিউল ১৬ পর্যন্ত যা যা আলোচনা হয়েছে, সেটার উপর ভিত্তি করে থাকবে একটা বড়সড় এসাইনমেন্ট ও লাইভ টেস্ট। ঝালিয়ে নিতে পারবেন নিজেকে।

Exam Week - 3

Containers, Kubernetes and Platform (Module 12-17)

Live Class 01: Shift-Left Security in CI Topics: SAST and code quality with SonarQube | Dependency and container scanning with Trivy | Secret scanning with Gitleaks | SBOM generation with Syft and image signing with Cosign | AI-assisted vulnerability triage: prioritize, explain and open a fix PR Live Class 02: Runtime Secrets and Policy as Code Topics: Secrets management with AWS Secrets Manager and External Secrets Operator | Policy as Code with Kyverno (or OPA Gatekeeper) | Blocking unsigned images and privileged pods at admission | Snyk / GitHub Advanced Security, automated security gate on every push Goal: A CI/CD pipeline and cluster where insecure or unsigned code cannot reach production. Project: Secure, Reliable and Cost-Smart Production Assignment: Show a pipeline rejecting a vulnerable image and the cluster rejecting an unsigned one.

DevSecOps and Software Supply Chain Security

Live Class 01: Reliability Engineering: SLOs, Alerts and Load Testing Topics: SLIs, SLOs and error budgets for the three-tier application | Alerting on user-facing symptoms instead of causes | Load testing with k6 | CloudWatch metrics, logs and alarms, Container Insights for EKS Live Class 02: AIOps: Incident Response with AI Agents Topics: CloudWatch anomaly detection | AWS DevOps Agent for investigation and root cause analysis | AI-generated incident timeline and postmortem draft | Human in the loop: when to trust the agent and when to escalate | Game day: inject a failure, resolve it with AI assistance, write a blameless postmortem Goal: Operate the application against defined SLOs and resolve incidents faster with AI, while keeping human accountability. Project: Secure, Reliable and Cost-Smart Production Assignment: Game day postmortem with SLO impact and an evaluation of the AI agent's findings.

SRE and AI-Powered Operations

Live Class 01: Serverless Architecture in Practice Topics: How serverless works, common use cases, key providers | Lambda, API Gateway and event-driven patterns (SQS, EventBridge) | Deploying serverless with Terraform | Pros and cons, and why some teams move back from serverless Live Class 02: FinOps: Cost as an Engineering Metric Topics: Cost allocation tags and AWS Budgets | Cost Explorer, Savings Plans, Reserved and Spot instances | Kubernetes cost visibility with OpenCost | Infracost: cost estimate on every Terraform pull request | Right-sizing with AI recommendations Goal: Deploy a serverless component and make cost visible and reviewable in every infrastructure change. Project: Secure, Reliable and Cost-Smart Production Assignment: A cost report for the full stack with three right-sizing changes applied.

Serverless Architecture and FinOps

Live Class 01: How AI Agents Work in Operations Topics: Assistant vs agent: autocomplete, chat and autonomous agents | Terminal coding agents (for example Claude Code or GitHub Copilot's agent) inside an ops workflow | Model Context Protocol (MCP): connecting agents to AWS, Kubernetes and GitHub | Agent writes Terraform and Helm changes as a pull request | Reviewing agent output: plan diff, policy checks, tests Live Class 02: Guardrails, Permissions and Governance for Agents Topics: Least-privilege IAM roles for agents (read-only vs write) | Human approval gates in CI/CD | Audit logging of agent actions with CloudTrail | Risks: prompt injection, secret leakage, runaway cost, destructive commands | Writing a team agent policy | Lab: an agent fixes a broken deployment through a PR that must pass policy and human review Goal: Use AI agents to make real infrastructure changes safely, with permissions, review gates and audit trails. Project: Add AI to the App, Safely Assignment: Agent policy document plus a merged, agent-authored PR that passed all gates.

Agentic DevOps: Operating Infrastructure with AI Agents

Live Class 01: Architecture and Deployment of an LLM Application Topics: Anatomy of an LLM / RAG application: API, model, vector store, prompts | Managed model APIs (Amazon Bedrock) vs self-hosted models (Ollama, vLLM) | pgvector on the course's existing PostgreSQL database | Containerizing the AI service and shipping it with Helm and Argo CD | Model API keys as managed secrets, rate limiting and timeouts Live Class 02: Running AI Workloads in Production Topics: GPU basics on Kubernetes: GPU node groups, NVIDIA device plugin, autoscaling (instructor demo, cost-aware) | LLM observability: latency, token usage, cost per request, error rates (OpenTelemetry GenAI conventions, Langfuse) | Model and prompt versioning with rollback | Evaluations as a deployment gate | Cost controls: caching, model routing, token budgets Goal: Add an AI feature to the three-tier application and run it in production with observability and cost controls. Project: Add AI to the App, Safely Assignment: An LLM dashboard showing latency, tokens and cost per request, plus a failing evaluation that blocks a deploy.

Deploying and Operating AI Applications (LLMOps)

মডিউল ১৮ থেকে মডিউল ২২ পর্যন্ত যা যা আলোচনা হয়েছে, সেটার উপর ভিত্তি করে থাকবে একটা বড়সড় এসাইনমেন্ট ও লাইভ টেস্ট। ঝালিয়ে নিতে পারবেন নিজেকে।

Exam Week - 4

Secure, Reliable and AI-Native Operations (Module 18-23)

Live Class 01: IAM, Organizations and ResilienceTopics: IAM roles and policies in depth, AWS Organizations and SCPs | S3 storage classes and lifecycle policies | Backup and disaster recovery strategies, CloudTrail logging Live Class 02: Well-Architected Review and Certification PreparationTopics: AWS Well-Architected Framework, six pillars | Running a Well-Architected review on the learner's own application | CloudFormation basics (brief, for exam coverage) | AWS Certified Solutions Architect Associate exam strategy Goal: Certification-ready AWS knowledge applied to the learner's own architecture. Project: Capstone: Production-Ready AI App, End to End Assignment: Written Well-Architected review with prioritized fixes.

AWS Deep Dive and Well-Architected Review

Live Class 01: Capstone Kickoff and Architecture ReviewTopics: Capstone brief and rubric walkthrough | Each team presents its architecture; instructors review risks and scope Live Class 02: Mid-Point Review and Unblock ClinicTopics: Progress demos, pipeline and cluster review | Live troubleshooting of blockers Goal: Capstone infrastructure, delivery and observability fully working. Project: Capstone: Production-Ready AI App, End to End Assignment: Capstone repository at mid-point milestone.

Capstone Build

Live Class 01: Capstone Demo DayTopics: Team demos, live failure injection and recovery | Panel feedback and scoring Live Class 02: 2026 Job Market and Interview PreparationTopics: Targeting roles: DevOps, Platform, SRE, DevSecOps, AI Infrastructure | Turning the capstone into a portfolio and LinkedIn case study | AI-era interview questions: reviewing generated IaC, agent governance, incident judgment | Mock technical interview Goal: A portfolio-ready capstone and readiness for 2026-style interviews. Project: Capstone: Production-Ready AI App, End to End Assignment: Final capstone submission and portfolio write-up.