333 hands-on projects across 111 courses: two Intermediate and one Advanced per course, each based on a real company use case. Read every brief here; SOTT students build them with step-by-step tasks and a trainer review.
Find, size and prioritise AI agent use cases for a sample company.
Build a working agent prototype for one prioritised use case.
Design the full architecture, governance, measurement and rollout for an enterprise agent programme.
Train, evaluate and explain a classifier for a business problem.
Train a small neural network to classify images and deploy it as an API.
Build a reproducible ML system with experiment tracking, a model registry and a deployed service.
Build an LLM agent that calls tools (search, calculator, database) to answer questions.
Build a helpdesk agent that resolves simple requests, remembers context and hands others to humans.
Build a multi-agent system (planner, researcher, writer, reviewer) with evaluation, guardrails and tracing.
Build a chatbot that answers common questions and hands over to a human.
Use Rekognition and Comprehend to tag images and analyse reviews.
Build a grounded assistant with Bedrock and explain its responsible use.
Extract data from forms with Document Intelligence and analyse text with Language.
Build two small apps: image captioning and speech-to-text notes.
Build a grounded assistant with Azure OpenAI and content safety.
Build a reusable prompt kit for your own daily work with ChatGPT, Claude, Gemini and Copilot.
Produce a short business report with data analysis, charts and slides made with AI help.
Design an AI-assisted workflow for a team process and a simple AI usage policy.
Write Playwright tests for a web app, using AI assistants to draft and refactor them.
Test REST APIs with Playwright request contexts, schemas and test data.
Build a combined UI and API framework with CI, parallel runs, visual checks and AI-generated test ideas.
Train, evaluate and deploy a churn model with Amazon SageMaker.
Extract, translate and analyse text from scanned documents with managed AI services.
Automate training, approval, deployment and drift monitoring with SageMaker Pipelines.
Host a personal or business website on cloud storage with a CDN and HTTPS.
Create a VM, secure it with identity and network rules, and monitor it.
Design a simple cloud architecture, cost estimate and migration plan for a small company.
Monitor EC2 and RDS with CloudWatch, alarms and automatic recovery actions.
Automate patching, backups and configuration compliance across servers.
Run a production-like environment with auto scaling, multi-AZ, security monitoring and cost controls.
Ingest raw files, transform them and load a reporting table on a schedule.
Stream click or delivery events, process them and show near real-time results.
Combine batch and streaming with governance, CI/CD and cost control.
Build a CRUD API with API Gateway, Lambda and DynamoDB.
Process orders with SQS, SNS and Step Functions with retries and a dead-letter queue.
Add authentication, tracing, CI/CD and safe deployments to a serverless app.
Build a pipeline that tests, builds and deploys a web app to Amazon ECS on every commit.
Describe a complete VPC, load balancer, auto scaling group and database in code and deploy it repeatably.
Run a microservice on EKS or ECS with blue/green releases, full observability and automatic rollback.
Build a question-answering assistant over company documents with Bedrock Knowledge Bases.
Create a Bedrock agent that uses action groups to look up and update orders.
Ship a Bedrock app with guardrails, evaluation, caching, cost tracking and monitoring.
Design and build a fault-tolerant web app with ALB, Auto Scaling and RDS Multi-AZ.
Build an upload-and-resize service with S3, Lambda, API Gateway and DynamoDB.
Review an existing app against the AWS Well-Architected Framework and redesign it for cost, security and reliability.
Set up identity, logging, guardrails and detection for a new account.
Protect data with KMS and Macie and networks with security groups, WAF and private endpoints.
Detect, respond and report on a simulated compromise with automation.
Set up identities, RBAC, policies, tags and budgets for a new subscription.
Build a hub-and-spoke VNet design with peering, NSGs, a firewall and a site-to-site VPN.
Run a VM and App Service workload with storage, backup, site recovery and monitoring.
Design management groups, policies, RBAC and identity for a company.
Design storage, database and app hosting for a web application with trade-offs.
Design backup, disaster recovery, high availability and a migration plan.
Connect Azure Boards, Repos and multi-stage Pipelines so every release is traceable to work items.
Speed up and standardise pipelines with templates, parallel jobs and caching.
Deliver a containerised app to AKS with security scanning, Key Vault, feature flags and monitoring.
Build a grounded chat app with Azure OpenAI and Azure AI Search.
Create an agent in Azure AI Foundry Agent Service that uses functions and code interpreter.
Deploy an AI app with content safety, evaluation, monitoring and a secure API.
Protect identities with Conditional Access, PIM and managed identities.
Protect workloads with NSGs, private endpoints, Key Vault and Defender for Cloud.
Secure an Azure OpenAI app end to end: identity, network, data, content safety and monitoring.
Design containers, partition keys and queries for a shopping cart API.
React to new orders with the change feed and update a materialised view.
Run a multi-region app with consistency choices, vector search and monitoring.
Load raw sales files into bronze, silver and gold Delta tables with pipelines and notebooks.
Build a T-SQL warehouse with dimensions and facts and a Power BI report on top.
Stream order events into an eventhouse, query with KQL, alert on problems and deploy with Git.
Ingest raw files, transform them and load a reporting table on a schedule.
Stream click or delivery events, process them and show near real-time results.
Combine batch and streaming with governance, CI/CD and cost control.
Build, test and deploy a web app to Azure App Service with multi-stage YAML pipelines.
Create a complete Azure environment from code, run from a pipeline with plan and approval steps.
Ship a microservice app to AKS with templates, security scanning, Key Vault secrets and monitoring.
Train models with Azure ML jobs and track experiments with MLflow.
Deploy a model to a managed endpoint and automate it with GitHub Actions.
Build pipelines, monitoring and retraining, plus evaluation for a generative AI flow.
Build a hub-and-spoke VNet design with peering, routing and a firewall.
Connect a simulated on-premises network to Azure with a site-to-site VPN and DNS.
Deliver a web app globally with Front Door, WAF, DDoS protection and monitoring.
Build a product database with stored procedures, views and JSON support.
Store embeddings in Azure SQL and search products by meaning.
Build a RAG assistant that answers from data in Azure SQL, with security and monitoring.
Deploy an Azure Virtual Desktop host pool with user profiles and app groups.
Build a golden image and a scaling plan to cut costs.
Add security, Conditional Access, monitoring, updates and disaster recovery.
Build a small REST API with routes, validation and JSON responses.
Connect the API to a database with models and queries.
Add sign-up, login with tokens, tests and a Docker image.
Call OpenAI, Claude and Gemini APIs from one small app and compare them.
Turn emails or invoices into validated JSON using structured outputs and tool calling.
Ship an LLM feature with streaming, prompt caching, rate limiting, evaluations and cost tracking.
Design and build a small office network with VLANs, DHCP, inter-VLAN routing and ACLs.
Connect head office and two branches with OSPF, redundancy and NAT to the internet.
Design a campus network with wireless, security and automation basics, then troubleshoot planted faults.
Design and build a redundant campus with EtherChannel, spanning tree tuning and first-hop redundancy.
Build multi-area OSPF and eBGP to two providers with route control.
Add security, wireless, SD-WAN concepts and automation with Python or Ansible.
Deploy a frontend, API and database on Kubernetes with config, storage, probes and scaling.
Upgrade a kubeadm cluster, back up and restore etcd, and manage access with RBAC.
Diagnose and fix broken nodes, networking, DNS, storage and scheduling, then write a troubleshooting guide.
Release a three-service app with rolling updates, a canary and one-command rollback.
Run reports and data jobs with Jobs, CronJobs and init containers, with retries and limits.
Make a full web app production-ready with probes, config, secrets, security context, Ingress with TLS and network policies.
Harden a kubeadm cluster to the CIS benchmark with audit logging and locked-down kubelets.
Scan, generate SBOMs for, sign and admit only trusted images, and minimise microservice vulnerabilities.
Detect a simulated attack with Falco and audit logs, then isolate, investigate and rebuild.
Create servers with Terraform and configure them with Ansible in one automated flow.
Containerise an app and deploy it to Kubernetes from a Jenkins pipeline.
Run dev, staging and production from Terraform modules and GitOps, and prove recovery in a second region.
Publish a simple web page from cloud storage and share the link.
Launch a virtual server, connect to it and serve a page.
Combine storage, a server and a managed database into a small working app.
Publish a simple web page from cloud storage and share the link.
Launch a virtual server, connect to it and serve a page.
Combine storage, a server and a managed database into a small working app.
Automate build, test and Docker image creation for a Java or Python app with Jenkins.
Provision a cluster with Terraform and deploy a multi-tier app with Helm.
Combine Terraform, Jenkins or GitHub Actions, Argo CD, Prometheus and Grafana into one working delivery platform.
Design and test roles, policies and conditional access across both clouds.
Encrypt data at rest and in transit with managed keys and detect sensitive data.
Build a secure landing zone for AWS and Azure with central logging, guardrails, networking and monitoring.
Build a responsive portfolio page with HTML and CSS.
Add JavaScript to make a to-do list that remembers items.
Build a multi-page site with a validated contact form and a data fetch.
Write a Dockerfile, build an image and run it with Docker and Podman.
Run an app and database together with networks and volumes.
Optimise and secure an image with multi-stage builds, non-root users and scanning.
Build a small lab with Kali and vulnerable machines, scan them and report the findings.
Segment a small network with VLANs and firewall rules and verify the segmentation.
Simulate a phishing-to-malware incident in a lab, detect it, respond and write the incident report.
Set up a new cloud account with secure identity, logging and storage defaults.
Turn on threat detection and automatically respond to a suspicious event.
Assess an account against CIS, fix the top risks, automate checks and report to management.
Clean sales data and build a management dashboard with KPIs and drill-down.
Answer business questions about customers and orders with SQL, including cohorts and RFM.
Take a messy business dataset from question to recommendation with Python, SQL and Power BI.
Use Copilot and AI features to clean data, write formulas and build a sales report faster.
Analyse customers with SQL and pandas, using an AI assistant to speed up code while checking every result.
Build a simple assistant that answers business questions from a sales database, with guardrails and checks.
Build a daily ETL pipeline that cleans raw files with PySpark and is orchestrated by Airflow.
Build bronze, silver and gold Delta tables on Databricks with Unity Catalog permissions.
Combine streaming events and batch data into one platform with tests, CI/CD and monitoring.
Clean, explore and explain a public dataset with Python.
Build and evaluate a model that predicts house prices or customer churn.
Frame a business question, build a model and share it as a simple web app.
Model, create and load a normalised database and answer business questions with SQL.
Protect and tune the database: backups, restore drills, indexes and query plans.
Run the database with high availability, security, automation and a disaster recovery plan.
Model, create and load a normalised database and answer business questions with SQL.
Protect and tune the database: backups, restore drills, indexes and query plans.
Run the database with high availability, security, automation and a disaster recovery plan.
Use Git and GitHub the way real teams do: branches, pull requests and reviews.
Run the app tests automatically on every push and pull request.
Build a Docker image in CI and run it, showing the full DevOps loop.
Turn a Java web app into an automated build that tests, packages and stores every version.
Configure web servers with Ansible roles and run the app in Docker containers.
Join Git, Jenkins, Docker, Kubernetes, Ansible, Prometheus and Grafana into one delivery chain.
Set up a branching strategy with protections, reviews and release tags.
Build a multi-stage pipeline with build, test and quality stages.
Extend the pipeline to deploy to a test server or container platform with approval.
Add code, dependency and secret scanning to a CI pipeline with quality gates.
Scan images and Terraform, harden a Dockerfile and use OIDC instead of stored cloud keys.
Build an end-to-end secure delivery chain: SBOMs, signing, admission control, DAST and runtime alerts.
Audit a website and plan content that brings the right visitors.
Plan and set up a Google or Meta ads campaign with conversion tracking.
Build a funnel from awareness to repeat purchase with content, ads, email and reporting.
Tag resources, build cost reports and show spend per team.
Find idle and oversized resources and calculate savings.
Set up budgets, commitments, unit economics and a monthly FinOps review.
Run a discovery with a "customer", map their workflow and design an AI solution with success metrics.
Connect an AI service to a "customer" system through APIs, webhooks and SSO.
Take an AI solution to production at a "customer": deployment, monitoring, evaluation and a value report.
Build a RAG app that answers questions over company PDFs with citations.
Measure RAG quality with a test set and improve it with hybrid search and re-ranking.
Deploy a RAG API with access control, prompt-injection defences, caching, cost tracking and monitoring.
Create a cloud VM, install a web server and deploy a site using only scripts.
Back up a database every night to cloud storage with Python, with alerts and a tested restore.
Build a load-balanced, auto-scaling web app with Terraform and a simple deployment pipeline.
Set up projects, identity, networking and a first workload with the right Google Cloud services.
Run a service with messaging, monitoring, logging and alerting.
Design and build a secure, reliable, cost-aware solution and review it against the Google Cloud Architecture Framework.
Ingest raw files, transform them and load a reporting table on a schedule.
Stream click or delivery events, process them and show near real-time results.
Combine batch and streaming with governance, CI/CD and cost control.
Model, create and load a normalised database and answer business questions with SQL.
Protect and tune the database: backups, restore drills, indexes and query plans.
Run the database with high availability, security, automation and a disaster recovery plan.
Set up projects, identity, networking and a first workload with the right Google Cloud services.
Run a service with messaging, monitoring, logging and alerting.
Design and build a secure, reliable, cost-aware solution and review it against the Google Cloud Architecture Framework.
Set up projects, identity, networking and a first workload with the right Google Cloud services.
Run a service with messaging, monitoring, logging and alerting.
Design and build a secure, reliable, cost-aware solution and review it against the Google Cloud Architecture Framework.
Set up projects, identity, networking and a first workload with the right Google Cloud services.
Run a service with messaging, monitoring, logging and alerting.
Design and build a secure, reliable, cost-aware solution and review it against the Google Cloud Architecture Framework.
Host a personal or business website on cloud storage with a CDN and HTTPS.
Create a VM, secure it with identity and network rules, and monitor it.
Design a simple cloud architecture, cost estimate and migration plan for a small company.
Train, evaluate and deploy a model on Vertex AI.
Build a grounded Gemini assistant over company documents.
Automate training and deployment with Vertex AI Pipelines and monitor the model.
Set up projects, identity, networking and a first workload with the right Google Cloud services.
Run a service with messaging, monitoring, logging and alerting.
Design and build a secure, reliable, cost-aware solution and review it against the Google Cloud Architecture Framework.
Set up identity, policy and data controls the way a security engineer would on day one.
Collect logs, detect suspicious activity and investigate it.
Combine prevention, detection, automation and reporting for a production environment.
Set up identity, policy and data controls the way a security engineer would on day one.
Collect logs, detect suspicious activity and investigate it.
Combine prevention, detection, automation and reporting for a production environment.
Design and run incident and service request processes in a service desk tool.
Run root cause analysis and a change approval process.
Measure services, find improvements and present a continual improvement plan.
Secure your own accounts and devices and explain each step.
Analyse sample emails and explain the warning signs.
Write a basic security plan for a small office: risks, controls and an incident checklist.
Build a REST API with Spring Boot, JPA and validation, with tests.
Add login, roles and a simple web UI to a claims application.
Split an order system into services that talk over REST and a message queue, run in Docker and deploy from CI.
Build a two-tier app with Docker and Compose, then move it to Kubernetes Deployments and Services.
Package an app as a Helm chart and let Argo CD deploy it from Git.
Add metrics, dashboards and alerts to an app on Kubernetes with Prometheus and Grafana.
Secure a team namespace with RBAC, Pod Security Admission, network policies and encrypted Secrets.
Audit a cluster with kube-bench and Kubescape and fix the most important findings.
Scan, generate SBOMs for and sign images in CI, and let only signed images into the cluster.
Deploy an app with a Deployment and Service on a local cluster.
Add ConfigMaps, Secrets, a volume and a rolling update.
Run a frontend, API and database with probes, scaling and Ingress.
Run a microservice app on a GitOps-managed cluster with Helm, Prometheus and Grafana.
Upgrade, back up and restore a kubeadm cluster and fix five broken scenarios.
Harden a cluster, sign and admit only trusted images, and detect attacks at runtime.
Set up a Linux server with users, packages, a web server and a firewall.
Write scripts for backups, disk checks and log clean-up, scheduled with cron.
Harden, patch, monitor and troubleshoot a production-style server.
Instrument an LLM app with tracing, token counts and cost per request.
Build an evaluation set and automated checks that run on every prompt change.
Run an LLM service with guardrails, online monitoring, A/B prompt rollout and alerts.
Set up a Microsoft 365 tenant with users, groups, licences and admin roles.
Apply Defender, DLP, retention and secure score improvements.
Combine identity, devices, Teams, SharePoint and governance into a rollout plan with a pilot.
Build and deploy a web app with managed identity and Key Vault.
Process uploads and messages with Azure Functions, Service Bus and Event Grid.
Build an app with containers, Cosmos DB, Azure OpenAI and API Management.
Create an Azure SQL database and a Cosmos DB container and compare them.
Load sample data and build a simple analytics report.
Combine storage, a simple pipeline, a database and a report for one business question.
Model, create and load a normalised database and answer business questions with SQL.
Protect and tune the database: backups, restore drills, indexes and query plans.
Run the database with high availability, security, automation and a disaster recovery plan.
Host a personal or business website on cloud storage with a CDN and HTTPS.
Create a VM, secure it with identity and network rules, and monitor it.
Design a simple cloud architecture, cost estimate and migration plan for a small company.
Enrol Windows devices with Autopilot and apply configuration profiles.
Set compliance policies, Windows updates and conditional access for devices.
Add Defender for Endpoint, app protection, remote actions and analytics across Windows and mobile.
Protect a tenant with MFA, Conditional Access and sign-in monitoring.
Automate access with groups, entitlement management and access reviews.
Add PIM, workload identities, app registrations, B2B and identity protection with reporting.
Create labels, auto-labelling and encryption for sensitive documents.
Stop sensitive data leaving through email, Teams, SharePoint and endpoints.
Combine retention, records, insider risk and eDiscovery into one programme.
Review identity, devices and data settings in a Microsoft 365 test tenant against zero trust principles.
Explore Compliance Manager, sensitivity labels and retention in a test tenant.
See how threats are detected and handled across Microsoft security tools.
Connect data sources, write KQL detections and triage incidents.
Investigate a simulated phishing-to-endpoint incident in Microsoft Defender XDR.
Automate enrichment and response with Sentinel playbooks and measure SOC performance.
Train models with MLflow tracking, compare runs and register the best model.
Serve a model as a FastAPI service in Docker and deploy it with a CI/CD pipeline.
Build an end-to-end MLOps pipeline: training, validation, deployment, drift monitoring and retraining.
Build a web, app and database tier on AWS and on Azure and compare the services.
Connect AWS and Azure with a site-to-site VPN and single sign-on.
Run an app on AWS with a warm standby on Azure, test fail-over, and add tagging, budgets and a cost dashboard.
Monitor routers, switches and servers with SNMP and ICMP and build a NOC dashboard.
Handle simulated network incidents with tickets, priorities, escalation and closure notes.
Build monitoring, a change process, config backups and automated checks into a NOC runbook.
Build a PHP app with forms, validation, sessions and MySQL.
Build a course catalogue with Laravel: routes, controllers, Eloquent and Blade.
Build a REST API with Sanctum auth, feature tests and deploy it to a server.
Give developers self-service namespaces and cloud resources through approved Terraform modules.
Create a Backstage software template that scaffolds a service with CI, Helm and docs.
Combine self-service, GitOps, policy as code, observability and cost labels into one platform.
Clean data, build a model and design a one-page sales dashboard.
Rebuild the analysis in Tableau and compare the tools.
Build a shared semantic model with security, refresh and documentation.
Find and define a real user problem with interviews and a problem statement.
Generate solutions, prioritise them and build a simple roadmap.
Write a short PRD with metrics and a launch plan for one feature.
Write a small console game with loops, conditions and functions.
Write a script that sorts files into folders by type.
Read sales data, calculate totals and write a summary report.
Write a charter, scope, schedule and budget for a realistic project.
Plan risks, stakeholders and communications for the same project.
Run a simulated project mixing predictive and agile, track performance with earned value and prepare for PMP.
Automate a monthly Excel report with Python, pandas and openpyxl.
Build a task-tracking API with FastAPI, SQLAlchemy, authentication and tests.
Build a tool that checks many servers or websites, stores results, alerts on problems and shows a dashboard.
Configure objects, fields, page layouts and a sales process for a small company.
Set up profiles, roles, sharing and a sales dashboard.
Build cases, queues, SLAs and Flow automation for customer support.
Configure objects, fields, page layouts and a sales process for a small company.
Set up profiles, roles, sharing and a sales dashboard.
Build cases, queues, SLAs and Flow automation for customer support.
Write an Apex trigger with a handler class and tests.
Build an LWC that shows and updates related records.
Connect Salesforce to an external REST API with callouts, platform events and CI.
Collect Windows and Linux logs in a SIEM and write rules for common attacks.
Investigate a phishing email end to end and write a reusable playbook.
Simulate a ransomware attack chain in a lab, detect each stage, respond and report.
Monitor latency, traffic, errors and saturation with Prometheus and Grafana.
Add distributed tracing and central logs to find where time is spent across services.
Define SLOs, alert on burn rate, run a game day and write a blameless post-mortem.
Write a versioned VPC module with remote state and use it for dev and prod.
Deploy the same web app to AWS and Azure from one Terraform repository.
Run Terraform through pull requests with plans, policy checks, tests and approvals.
Write a test plan, test cases and bug reports for a web app.
Test a REST API with collections, assertions and environments.
Build a maintainable UI automation framework with page objects, reports and CI.
Interview users, build personas and map a user journey.
Turn findings into wireframes and an interactive Figma prototype.
Test the prototype with users, improve it and write a portfolio case study.
Interview users, build personas and map a user journey.
Turn findings into wireframes and an interactive Figma prototype.
Test the prototype with users, improve it and write a portfolio case study.
Interview users, build personas and map a user journey.
Turn findings into wireframes and an interactive Figma prototype.
Test the prototype with users, improve it and write a portfolio case study.
Build a domain with users, groups, OUs and Group Policy.
Set up file shares, permissions, quotas and backup.
Manage on-premises servers with Azure Arc, update management, monitoring and security.
Pick projects that match the job you want and show real tools end to end: a CI/CD pipeline and infrastructure as code for DevOps roles, a Kubernetes platform for Kubernetes and platform roles, a data pipeline for data roles, or a RAG app for AI roles. One Advanced project with a clear README, diagram and demo video is worth more than five copied tutorials.
The project briefs on this page are free to read. Enrolled SOTT students get the step-by-step tasks, "done when" checklists, a place to submit and a review by their trainer in the SOTT student portal.
Intermediate projects build one working system with the core tools of the course. Advanced projects combine several systems the way a real company does, with security, monitoring, automation and a written runbook or report.
Yes, for enrolled students. You submit a repository link and a report or screenshots, and your trainer approves it or asks for specific changes. Approved projects are ready for your resume and LinkedIn.