2366 fiches de synthèse

• Networking Ports and ProtocolsFondamentaux

  Fundamentals | Part 1: ports every SOC analyst and exam candidate must know, secure vs insecure protocols, exposure risk (part 2: OSI/TCP-IP layers, TCP flags, nmap scans, encapsulation)

  Télécharger (PDF)

• IPv4 Addressing, Header and SubnettingRéseaux

  Networking track, part 1 of IPv4 | Binary and masks, the IPv4 header field by field, mask arithmetic with worked answers, classful legacy (VLSM, summarization, special ranges and NAT are in part 2) | Sources: Tanenbaum, Computer Networks 5th ed. 5.6; RFC 791, 1918, 3021, 4632

  Télécharger (PDF)

• Zero-Trust ArchitectureArchitectures de référence

  The NIST SP 800-207 logical model as a diagram: Policy Engine, Policy Administrator and Policy Enforcement Point; the subject-to-resource request flow and the inputs a decision uses; the pillars, ZTNA vs VPN, and migration from the perimeter model | maturity levels reference 'CISA Zero Trust Maturity Model v2.0'; tenets and trust algorithm are detailed in 'Zero Trust Architecture SP 800-207'

  Télécharger (PDF)

• CPUs, Memory, Motherboards and BusesMatériel et racines de confiance

  CompTIA A+ 220-1201 domain 3.0 and beyond | x86/ARM, sockets, DDR4/DDR5, ECC, PCIe, USB4 | CPU security features (ANSSI DAT-24)

  Télécharger (PDF)

• Wireshark and tshark for Troubleshooting and AnalysisOutils

  Tools track | Capture vs display filters, per-protocol filters, follow stream, statistics and IO graphs, decrypting your own TLS, tshark field extraction, colouring rules and profiles

  Télécharger (PDF)

• Writing for Security WorkPratique professionnelle

  Professional track | Incident reports, findings, executive summaries, tickets, runbooks and playbooks, post-incident reviews: templates, plain-language rules, evidence and timestamps, severity and certainty wording, audience adaptation

  Télécharger (PDF)

• SIEM and Detection Engineering (part 1)Défense (blue team)

  Blue team track | What a SIEM does, the detection engineering lifecycle, the log pipeline from host to alert, metrics that matter (part 2: Sigma rules, coverage, correlation and triage)

  Télécharger (PDF)

• OWASP Top 10 - Web (2021 and 2025)OWASP et sécurité applicative

  OWASP Top 10:2021 (A01-A10) explained with CWE names, detection coverage, 2017 and 2025 deltas, secure configuration

  Télécharger (PDF)

• Kubernetes SecurityCloud et DevSecOps

  4C model, threat model, Pod Security Standards, RBAC, NetworkPolicy, admission control (part 2: secrets, runtime, response) | CKS, CCSP, SY0-701

  Télécharger (PDF)

• Engagement Methodology and ReportingTests d'intrusion (cadre)

  CompTIA PenTest+ PT0-003 Domain 1.0 | EC-Council CEH v13 Module 01 | OffSec OSCP report | PTES, OSSTMM, NIST SP 800-115 (Technical Guide to Information Security Testing and Assessment), OWASP WSTG

  Télécharger (PDF)

• NIS2 Directive: EU-wide ObligationsGouvernance et réglementation UE

  Directive (EU) 2022/2555, in force 16 Jan 2023, transposition due 17 Oct 2024 | Scope, Art. 20-21 governance and measures, Art. 23 reporting; part 2: supervision, DORA / CER / CRA, ISO mapping, significance thresholds

  Télécharger (PDF)

• CompTIA CySA+ CS0-003Certifications CompTIA

  CompTIA Cybersecurity Analyst (CySA+) exam CS0-003 | Exam facts, domain weights, question formats, study plan, acronym drill (objectives decoded in part 2)

  Télécharger (PDF)

• Security Certification LandscapeCertifications avancées

  CISSP, CISM, CISA, CRISC, CCSP, CEH, OSCP and the CompTIA track | who issues what, exam format, experience rules, positioning (part 1 of 2; part 2: career paths, how each exam thinks, recent changes)

  Télécharger (PDF)

• NIST Cybersecurity Framework 2.0NIST

  NIST CSF 2.0 (26 Feb 2024) | Part 1: Core (Functions, Categories, Subcategories), Tiers, Profiles (part 2: risk communication, 1.1 changes, NIS2 and DORA mapping)

  Télécharger (PDF)

• IT Hygiene Guide - 42 Essential MeasuresANSSI

  ANSSI Guide d'hygiene informatique v2.0 (2017) | how the guide works, 10 themes and 42 measures, standard vs strengthened levels, effort and impact (core measures in practice and NIS2 mapping: part 2)

  Télécharger (PDF)

• NSA CSI: MCP Security Design Considerations (1/3)Référentiels et rapports

  NSA Cybersecurity Information Sheet U/OO/6030316-26, PP-26-1834, May 2026 v1.0 | What MCP is, why its design inverts the usual client-server pattern, and the nine security concerns the NSA observed in design, implementation and practice

  Télécharger (PDF)

• LLM App Security Testing 1: Threats and HarnessSujets émergents

  Defensive testing of LLM-based apps and agents, part 1 | Threat model, attack classes, test plan, guardrails, garak / promptfoo / PyRIT, OWASP LLM Top 10 2025 and MITRE ATLAS mapping

  Télécharger (PDF)

• Cyber GlossaryGlossaire

  7626 terms collected from the sheets in this collection; [brackets] name the source sheets

  Télécharger (PDF)

• AWS Big Picture for BeginnersAWS · Start here

  AWS overview, Overview of Amazon Web Services whitepaper, as of 2026 | What the cloud is, service models, deployment models, how an app maps to services, what AWS runs for you, control vs data plane, how AWS bills, service families

  Télécharger (PDF)

• Build Roles: Cloud Engineer, Architect, DeveloperAWS · Careers and jobs

  Careers chapter | Daily work, skills, tools, certifications, portfolio projects and interview questions with model answers for cloud engineers, solutions architects and cloud developers

  Télécharger (PDF)

• Atlas: Compute and ContainersAWS · Services atlas

  AWS Compute and Containers categories, as of 2026 | EC2 families and naming, Auto Scaling, Lightsail, Lambda, Batch, Elastic Beanstalk, App Runner, ECS, EKS, Fargate, ECR, Outposts and edge compute, HPC; when to choose each and the look-alikes

  Télécharger (PDF)

• CLF-C02 Domain 1: Cloud ConceptsAWS · Cloud Practitioner (CLF-C02)

  CLF-C02 exam guide Domain 1 (24 %), Overview of AWS whitepaper, Well-Architected Framework, AWS CAF, as of 2026 | Benefits and six advantages, design principles and the six pillars, migration (CAF, 7 Rs, tools), cloud economics (CapEx/OpEx, TCO, rightsizing, licensing)

  Télécharger (PDF)

• AIF-C01 Domain 1: Fundamentals of AI and MLAWS · AI Practitioner (AIF-C01)

  AIF-C01 exam guide Domain 1 (20 %), as of 2026 | 1.1 AI terms and learning types, 1.2 practical use cases and when not to use ML, AWS task AI services, 1.3 the ML lifecycle, SageMaker AI components, MLOps, model and business metrics

  Télécharger (PDF)

• AIB-C01 Domain 1: AI Fundamentals and LiteracyAWS · AI Business Strategist (AIB-C01)

  AIB-C01 exam guide Domain 1 (24 %), ISO/IEC 22989 and 23053 vocabulary, as of 2026 | 1.1 AI concepts and data for business readers; 1.2 rules vs AI, agents, drift, shadow-AI classification; 1.3 prompting, tokens and context windows, RAG vs fine-tuning

  Télécharger (PDF)

• SAA-C03 Secure 1: IAM, Organizations, EncryptionAWS · Solutions Architect Associate (SAA-C03)

  SAA-C03 Domain 1 Design Secure Architectures (30 %), tasks 1.1 secure access and 1.3 data security | IAM users, roles, policies, federation, Organizations and SCPs, KMS, Secrets Manager, ACM, threat-detection services

  Télécharger (PDF)

• DVA-C02 Domain 1: SDKs, Credentials and Lambda CodeAWS · Developer Associate (DVA-C02)

  AWS Certified Developer - Associate, Domain 1 (Development with AWS Services) part 1 | SDK credential chain, retries and backoff, pagination and waiters, Lambda handler model, invocation types, limits, idempotency

  Télécharger (PDF)

• SOA-C03 Domain 1: Monitoring, Logging, RemediationAWS · CloudOps Engineer Associate (SOA-C03)

  AWS Certified CloudOps Engineer Associate exam guide, as of 2026 | CloudWatch metrics, alarms, agent, Logs and Logs Insights, CloudTrail, EventBridge, Systems Manager Automation, Config remediation, Health

  Télécharger (PDF)

• DEA-C01 Domain 1a: Data Ingestion, Batch and StreamingAWS · Data Engineer Associate (DEA-C01)

  AWS DEA-C01 exam guide task 1.1, AWS service docs, as of 2026 | Kinesis Data Streams, Data Firehose, MSK, DMS, AppFlow, DataSync, Transfer Family, Snowball, S3 events and EventBridge; throughput limits, replay, fan-out, schedules and triggers

  Télécharger (PDF)

• MLA-C01 Domain 1: Data Preparation for MLAWS · Machine Learning Engineer Associate (MLA-C01, MLA-C02)

  AWS MLA-C01 exam guide tasks 1.1-1.3, SageMaker AI developer guide, as of 2026 | Ingest and store training data, formats and input modes, transform and engineer features (Data Wrangler in Canvas, Processing, Feature Store), labelling, data integrity, bias checks with Clarify, splits

  Télécharger (PDF)

• SAP-C02 Org Complexity 1: Multi-Account GovernanceAWS · Solutions Architect Professional (SAP-C02)

  SAP-C02 Domain 1 Design Solutions for Organizational Complexity (26 %), tasks 1.3-1.5 | OU design, Control Tower and account vending, SCPs, RCPs, declarative and tag policies, Identity Center and ABAC, delegated administration, centralized logging and security, cost visibility

  Télécharger (PDF)

• DOP-C02 Domain 1: SDLC Automation and PipelinesAWS · DevOps Engineer Professional (DOP-C02)

  AWS Certified DevOps Engineer - Professional, Domain 1 (22 %) | CodePipeline V2, CodeBuild, CodeArtifact, ECR, test stages, cross-account delivery; the DVA-C02 chapter teaches the basics

  Télécharger (PDF)

• AIP-C01: Foundation Models and Bedrock APIsAWS · Generative AI Developer Professional (AIP-C01)

  AIP-C01 Domain 1 (foundation model integration), Amazon Bedrock user guide and API reference, as of 2026 | Choosing a model, Converse vs InvokeModel, streaming, tool use, inference profiles and cross-region inference, throughput and service tiers, batch, prompt caching, quotas

  Télécharger (PDF)

• SCS-C03 Domain 1: Detection, Monitoring and LoggingAWS · Security Specialty (SCS)

  AWS Certified Security - Specialty SCS-C03 exam guide, Domain 1 (16 %) | GuardDuty, Security Hub, Inspector, Macie, Config, CloudTrail, VPC Flow Logs, Security Lake, CloudWatch; analysis and troubleshooting

  Télécharger (PDF)

• ANS-C01: Direct Connect, VPN and BGPAWS · Advanced Networking Specialty (ANS-C01)

  AWS Certified Advanced Networking Specialty exam guide, as of 2026 | Direct Connect connections and VIFs, Direct Connect gateway, resiliency models, Site-to-Site and Client VPN, BGP path control, MTU, MACsec

  Télécharger (PDF)

• Well-Architected: Operational Excellence and SecurityAWS · Well-Architected Framework

  AWS Well-Architected Operational Excellence and Security pillar whitepapers (as of 2026) | design principles, best-practice areas, what the OPS and SEC questions ask, AWS services per practice, metrics, anti-patterns

  Télécharger (PDF)

• Amazon ECS in DepthAWS · Containers and EKS

  Amazon ECS Developer Guide (as of 2026) | clusters, task definitions, services, launch types and capacity providers, network modes, Service Connect, load balancing, deployments, IAM roles, secrets, logging, storage, auto scaling, troubleshooting

  Télécharger (PDF)

• AWS Lambda in Depth: Execution, Invocation, FailureAWS · Serverless

  AWS Lambda developer guide, as of 2026 | execution environment lifecycle, runtimes and container images, synchronous / asynchronous / poll-based invocation, event source mappings, destinations and DLQs, permissions, layers and extensions

  Télécharger (PDF)

• CloudFormation Templates in DepthAWS · Infrastructure as code and CLI

  AWS CloudFormation user guide and template reference, as of 2026 | template sections, parameters and types, intrinsic functions, pseudo parameters, conditions and mappings, outputs and cross-stack exports, dynamic references, DeletionPolicy and update behaviours, quotas

  Télécharger (PDF)

• AWS Pricing Models and the BillAWS · Cost and FinOps

  AWS pricing pages, Billing and Cost Management user guide, as of 2026 | Pricing dimensions, On-Demand, Savings Plans, Reserved Instances, Spot, capacity reservations, Dedicated Hosts, free tier, data transfer, consolidated billing, how to read a bill

  Télécharger (PDF)

• Machine Learning Specialty (MLS-C01): LegacyAWS · Retired and legacy exams

  AWS Certified Machine Learning - Specialty exam guide (MLS-C01) and its retirement announcement, as of 2026 | What it tested, the ML science still worth knowing, where each domain lives now (MLA-C01, AIP-C01, DEA-C01), reading MLS on a CV

  Télécharger (PDF)

• Glossary - Advanced Networking Specialty Ans C01AWS · Glossary

  102 terms collected from the sheets in this collection; [brackets] name the source sheets

  Télécharger (PDF)

• Azure Global InfrastructureAZURE · Start here

  Azure regions, availability zones and reliability documentation, as of 2026 | Geographies, regions, region pairs, availability zones, datacenters, sovereign clouds, edge; choosing a region; SLAs and composite availability

  Télécharger (PDF)

• Infrastructure and Architecture RolesAZURE · Careers and jobs

  Microsoft Learn role descriptions (AZ-104, AZ-700, AZ-802, AZ-140, AZ-305), Cloud Adoption Framework roles, public job ads, as of 2026 | Cloud admin, cloud engineer, network engineer, hybrid admin, end-user computing, solutions architect, FinOps: tasks, tools, certifications, KPIs

  Télécharger (PDF)

• Atlas: Compute and ContainersAZURE · Services atlas

  Azure compute decision tree, Virtual Machines, App Service, Functions, Container Apps and AKS documentation, as of 2026 | Every compute service with purpose, when to choose, look-alikes, AWS and Google Cloud equivalents; VM families; scaling

  Télécharger (PDF)

• AZ-900 Domain 1: Cloud ConceptsAZURE · Azure Fundamentals (AZ-900)

  AZ-900 skills measured 2026-07-20, 'Describe cloud concepts' (25-30 %) | Cloud computing, shared responsibility, public / private / hybrid, consumption model and pricing models, serverless, benefits of the cloud, IaaS / PaaS / SaaS

  Télécharger (PDF)

• AI-901: Responsible AI and Generative ModelsAZURE · Azure AI Fundamentals (AI-901)

  AI-901 domain 1, Identify AI concepts and capabilities (40-45 %), skills measured as of April 15, 2026 | Six responsible AI principles, how generative models work, choosing a model, deployment options and configuration parameters

  Télécharger (PDF)

• AI-900: AI Workloads and Responsible AIAZURE · Azure AI Fundamentals (AI-900), retired 2026

  AI-900 domain 1, Describe AI workloads and considerations (15-20 %, as of 2026) | Workload types, matching problems to services, Microsoft's six responsible AI principles, tools that apply them

  Télécharger (PDF)

• DP-900: Core Data ConceptsAZURE · Azure Data Fundamentals (DP-900)

  DP-900 domain 1, Describe core data concepts (25-30 %, as of 2026) | Structured, semi-structured and unstructured data, file formats, databases, transactional vs analytical workloads, data roles and services

  Télécharger (PDF)

• SC-900: Security, Compliance and Identity ConceptsAZURE · Security, Compliance and Identity Fundamentals (SC-900)

  SC-900 domain 1, Describe the concepts of security, compliance and identity (10-15 %, as of 2026) | Shared responsibility, defence in depth, zero trust, encryption and hashing, GRC, authentication vs authorisation, identity providers, federation

  Télécharger (PDF)

• AZ-104: Entra ID Users, Groups and LicencesAZURE · Azure Administrator Associate (AZ-104)

  AZ-104 domain 1 (identities and governance), part 1 | Microsoft Entra ID tenants, users, groups, dynamic membership, licences, external (B2B) users, administrative units, SSPR, devices; az CLI, Microsoft Graph PowerShell

  Télécharger (PDF)

• AI-200 Containers 1: ACR, ACR Tasks, App ServiceAZURE · Azure AI Cloud Developer Associate (AI-200)

  AI-200 skills measured, domain 1 'Develop containerized solutions on Azure' (20-25 %), 'Implement container application hosting', as of 2026 | Build, store, version and manage images in Azure Container Registry; ACR Tasks; containers on App Service with environment variables and secrets

  Télécharger (PDF)

• AZ-204 Compute 1: Azure App Service Web AppsAZURE · Azure Developer Associate (AZ-204), retired 2026

  AZ-204 domain 1, Develop Azure compute solutions | Plans and tiers, deploy, configuration, deployment slots and swap, autoscale, TLS, diagnostics logs; az CLI; as of 2026

  Télécharger (PDF)

• SC-500 Domain 1: Entra Access, PIM and App IdentitiesAZURE · Cloud and AI Security Engineer Associate (SC-500)

  SC-500 skills measured: Secure access to resources by using Microsoft Entra ID | PIM, Conditional Access, MFA and passwordless, app registrations, OAuth consent, managed identities

  Télécharger (PDF)

• AZ-500 Domain 1: Secure Identity and AccessAZURE · Azure Security Engineer Associate (AZ-500), retired 2026

  AZ-500 skills measured, domain 1 (15-20 %), as of 2026 | Entra roles vs Azure RBAC, scopes, custom roles, conditions, PIM, managed identities, service principals, app permissions, MFA recall

  Télécharger (PDF)

• VNets, IP Address Planning, Public IPs and DNSAZURE · Azure Network Engineer Associate (AZ-700)

  AZ-700 domain 1, Design and implement core networking infrastructure (part 1) | Address spaces and subnets, reserved addresses, special subnets and delegation, public IP SKUs and prefixes, default outbound access, Azure DNS, Private DNS zones, DNS Private Resolver; az CLI and Bicep

  Télécharger (PDF)

• AZ-802: Domain Controllers, FSMO, Sites and TrustsAZURE · Windows Server Hybrid Administrator Associate (AZ-802)

  AZ-802 domain 1 (Deploy and manage AD DS, 20-25 %), part 1 | DCs on-premises and in Azure, RODCs and their account security, FSMO roles, trusts, sites and replication; PowerShell, ntdsutil, repadmin

  Télécharger (PDF)

• AZ-800: AD DS Forests, Domains, Sites and GPOsAZURE · Windows Server Hybrid Administrator (AZ-800, AZ-801), retired 2026

  AZ-800 domain 1 (AD DS on-premises and in the cloud), part 1 | Forests, domains, partitions, FSMO roles, functional levels, deploying DCs and RODCs, sites and replication, trusts, Group Policy, fine-grained password policies, recycle bin, gMSA; PowerShell and repadmin

  Télécharger (PDF)

• AI-103 D1: Choose Foundry Services and Set UpAZURE · Azure AI Apps and Agents Developer Associate (AI-103)

  AI-103 skills measured, Plan and manage an Azure AI solution (25-30 %), as of April 2026 | Choose models, Foundry services, retrieval, memory and tools; design the infrastructure; deployment types and configuration; Foundry projects in CI/CD

  Télécharger (PDF)

• AI-102 Domain 1: Plan and Manage Azure AIAZURE · Azure AI Engineer Associate (AI-102), retired 2026

  AI-102 skills measured, domain 1 (20-25 %), as of 2026 | Choosing services and resources, Foundry projects, authentication, networking, quotas, monitoring, cost, containers, Content Safety and responsible AI

  Télécharger (PDF)

• AI-300 Domain 1: MLOps Infrastructure as CodeAZURE · Machine Learning Operations (MLOps) Engineer Associate (AI-300)

  AI-300 skills measured, domain 1 'Design and implement an MLOps infrastructure' (15-20 %), as of 2026 | Workspace resources (datastores, compute, identity), assets (data, environments, components, registries), IaC with Bicep and Azure CLI, GitHub Actions with OIDC, network isolation, Git source control

  Télécharger (PDF)

• DP-100 Domain 1: Design and Prepare an ML SolutionAZURE · Azure Data Scientist Associate (DP-100), retired 2026

  DP-100 skills measured, domain 1 (20-25 %), as of 2026 | Workspace and its resources, compute targets, datastores and data assets, environments, registries, SDK v2 / CLI v2, security (RBAC, managed VNet, identities), Git and cost

  Télécharger (PDF)

• DP-300 Domain 1: Plan and Deploy Azure SQLAZURE · Azure Database Administrator Associate (DP-300)

  Microsoft Learn DP-300 skills outline and Azure SQL documentation, as of 2026 | Azure SQL Database vs Managed Instance vs SQL Server on Azure VMs, DTU and vCore, service tiers, serverless, elastic pools, sizing, deployment with az CLI and Bicep, migration paths

  Télécharger (PDF)

• DP-600 Maintain (1): Capacity, Security and GovernanceAZURE · Fabric Analytics Engineer Associate (DP-600)

  Microsoft Learn DP-600 skills outline and Microsoft Fabric documentation, as of 2026 | Tenant, capacity, domain, workspace and item hierarchy, F SKUs, smoothing and throttling, workspace roles, item permissions, OneLake security, RLS/CLS/OLS, dynamic data masking, sensitivity labels, endorsement, auditing

  Télécharger (PDF)

• DP-700 Domain 1: Implement and Manage the SolutionAZURE · Fabric Data Engineer Associate (DP-700)

  Microsoft Learn DP-700 skills outline and Microsoft Fabric documentation, as of 2026 | Workspace and Spark settings, pools and environments, domains, OneLake settings, security and governance, network protection, lifecycle (Git, deployment pipelines, database projects, variable libraries), orchestration with pipelines, notebooks and dataflows, parameters, schedules and event triggers

  Télécharger (PDF)

• DP-750 Domain 1a: Select and Configure ComputeAZURE · Azure Databricks Data Engineer Associate (DP-750)

  DP-750 study guide, 'Select and configure compute in a workspace'; Azure Databricks compute docs, as of 2026 | Compute types, access modes, sizing, autoscaling, termination, pools, Photon, runtimes, libraries, permissions, policies, cost

  Télécharger (PDF)

• DP-800 Domain 1a: Database Objects and Specialised TablesAZURE · SQL AI Developer Associate (DP-800)

  DP-800 study guide, 'Design and implement database objects'; SQL Server 2025, Azure SQL and Fabric SQL database docs, as of 2026 | Data types, indexes and columnstore, in-memory, temporal, external, ledger and graph tables, JSON columns and indexes, constraints, sequences, partitioning

  Télécharger (PDF)

• SC-200 Domain 1: The Security Operations EnvironmentAZURE · Security Operations Analyst Associate (SC-200)

  SC-200 skills measured, domain 1 Manage a security operations environment (20-25 %), as of 2026 | Defender portal and unified SecOps, unified RBAC, device groups and automation levels, alert tuning, indicators, advanced features, Sentinel onboarding, data lake and retention, Security Copilot set-up

  Télécharger (PDF)

• SC-300 Domain 1a: Tenant, Users, Groups and RolesAZURE · Identity and Access Administrator Associate (SC-300)

  SC-300 skills measured, domain 1 Implement and manage user identities (20-25 %), as of 2026 | Tenant settings and custom domains, users and licences, group types and dynamic membership, administrative units, Entra built-in and custom roles, Graph PowerShell

  Télécharger (PDF)

• Classify Data: SITs, EDM, Fingerprints, ClassifiersAZURE · Information Security Administrator Associate (SC-401)

  SC-401 domain 1, Implement information protection (part 1) | Sensitive information types, confidence and proximity, exact data match, document fingerprinting, trainable classifiers, content and activity explorer

  Télécharger (PDF)

• AZ-305 Domain 1a: Identity and Governance DesignAZURE · Azure Solutions Architect Expert (AZ-305)

  AZ-305 skills outline, domain 1 (identity, governance), as of 2026 | Entra ID tenant and external identities, authentication and authorization design, managed identities, Key Vault, management groups, Azure Policy, RBAC, tagging, locks

  Télécharger (PDF)

• AZ-400 Domain 1: Processes and CommunicationsAZURE · DevOps Engineer Expert (AZ-400)

  AZ-400 skills measured, domain 1 (about 10-15 %, as of 2026) | Work tracking in Azure Boards and GitHub, traceability, flow and DORA metrics, dashboards, wikis and documentation, notifications and chat integrations

  Télécharger (PDF)

• Zero Trust Strategy and Microsoft Security FrameworksAZURE · Cybersecurity Architect Expert (SC-100)

  SC-100 domain 1, Design solutions that align with security best practices and priorities (part 1) | Zero Trust principles and RaMP, MCRA, Microsoft cloud security benchmark, CAF Secure methodology, Well-Architected security pillar, translating business goals into a security strategy

  Télécharger (PDF)

• Plan AVD: Host Pools, App Groups, Capacity, LicensingAZURE · Azure Virtual Desktop Specialty (AZ-140)

  AZ-140 domain 1, Plan and implement an Azure Virtual Desktop infrastructure (part 1) | Architecture and object model, pooled vs personal, load balancing and session limits, sizing users per vCPU, regions and metadata, AVD vs Windows 365 vs Azure Local, licensing; PowerShell and az CLI

  Télécharger (PDF)

• AZ-120 Domain 1: Migrate SAP Workloads to AzureAZURE · Azure for SAP Workloads Specialty (AZ-120)

  AZ-120 skills outline (migrate SAP workloads), SAP on Azure planning guide, as of 2026 | Assessment and sizing (SAPS, HANA memory), migration methods (homogeneous, heterogeneous, DMO, HSR), data transfer, downtime planning, tooling and licensing

  Télécharger (PDF)

• DP-420 Domain 1a: Model Documents in Cosmos DBAZURE · Azure Cosmos DB Developer Specialty (DP-420)

  Microsoft Learn DP-420 skills outline and Azure Cosmos DB for NoSQL modelling documentation, as of 2026 | Embed vs reference, denormalisation, mixed entity types in one container, item size and ids, system properties, schema versioning, unique keys, TTL, modelling for read and write patterns

  Télécharger (PDF)

• CAF Strategy and PlanAZURE · Cloud Adoption and Well-Architected

  Microsoft Cloud Adoption Framework, Strategy and Plan methodologies, as of 2026 | Motivations and business outcomes, business case, digital estate and rationalisation, skills and organisation, the cloud adoption plan

  Télécharger (PDF)

• AKS Architecture and Cluster SetupAZURE · Containers and AKS

  Microsoft Learn AKS documentation, as of 2026 | Control plane and pricing tiers, Kubernetes version support, system and user node pools, VM sizes and OS, availability zones, AKS Automatic vs Standard, az CLI, Bicep and Terraform

  Télécharger (PDF)

• ARM Templates: Structure, Scopes and ModesAZURE · Infrastructure as code and CLI

  Microsoft Learn: ARM template reference, template functions, deployment modes, template specs, deployment scripts, as of 2026 | The JSON engine under Bicep: sections, expressions, four scopes, Incremental vs Complete, linked templates, template specs

  Télécharger (PDF)

• Azure Pricing and Purchase OptionsAZURE · Cost management and FinOps

  Azure pricing, reservations, savings plan for compute, Azure Hybrid Benefit, spot, dev/test and free offers, data transfer, as of 2026 | How each option bills, which applies first, break-even utilisation, estimating with the pricing and TCO calculators

  Télécharger (PDF)

• Retired Data and AI Exams: DP-203, DP-200/201, DA-100, DP-500, AI-100AZURE · Retired and legacy exams

  Microsoft Learn retirement notices and successor exam pages, as of 2026 | What each exam certified, what replaced it, which skills still apply, how to read the job ad

  Télécharger (PDF)

• Glossary - Administrator Associate Az 104AZURE · Glossary

  163 terms collected from the sheets in this collection; [brackets] name the source sheets

  Télécharger (PDF)

• Digital Transformation with Google CloudGCP · Cloud digital leader

  GCP Cloud Digital Leader | Cloud Concepts, Migration & Innovation

  Télécharger (PDF)

• Compute & NetworkingGCP · Associate cloud engineer

  GCP Associate Cloud Engineer | Compute Engine, GKE, Serverless, VPC & Load Balancing

  Télécharger (PDF)

• Compute & ContainersGCP · Professional cloud architect

  GCP Professional Cloud Architect | Compute Engine, GKE, Serverless & Choosing Compute

  Télécharger (PDF)

• VPC & Hybrid ConnectivityGCP · Professional cloud network engineer

  GCP Professional Cloud Network Engineer | VPC Architecture, Interconnect, VPN & Cloud Router

  Télécharger (PDF)

• Identity & Access ManagementGCP · Professional cloud security engineer

  GCP Professional Cloud Security Engineer | IAM, Workload Identity, Org Policies & BeyondCorp

  Télécharger (PDF)

• Data Ingestion & ProcessingGCP · Professional data engineer

  GCP Professional Data Engineer | Pub/Sub, Dataflow, Dataproc & Processing Patterns

  Télécharger (PDF)

• Production ML SystemsGCP · Professional ml engineer

  GCP Professional ML Engineer | Architecting, Adaptability, Performance & Hybrid

  Télécharger (PDF)

• Application DevelopmentGCP · Professional cloud developer

  GCP Professional Cloud Developer | Cloud Run, App Engine, Cloud Functions & Design Patterns

  Télécharger (PDF)

• CI/CD Pipelines & Supply Chain SecurityGCP · Professional cloud devops engineer

  GCP Professional Cloud DevOps Engineer | Cloud Build, Cloud Deploy, Binary Authorization & SLSA

  Télécharger (PDF)

• Relational DatabasesGCP · Professional database engineer

  GCP Professional Database Engineer | Cloud SQL, Cloud Spanner, AlloyDB & Schema Design

  Télécharger (PDF)

• Vertex AI for Generative AIGCP · Generative ai and agents

  GCP Generative AI | Model Garden, Gemini, Tuning, RAG & Responsible AI

  Télécharger (PDF)

• GKE ArchitectureGCP · Kubernetes and gke

  GCP Kubernetes & GKE | Cluster Modes, Node Pools, Networking & Security

  Télécharger (PDF)

• How to Use This AI CourseStart here

  Welcome | The 16 folders and what each one owns, four reading paths (practitioner, ML engineer, researcher, leader), how every sheet is built, a realistic pace and a spaced-review schedule

  Télécharger (PDF)

• Algebra and Notation Refresher for MLMaths

  Maths for AI | Sets, functions, sums Σ and products (capital Pi), exponents and logarithms, indices, and a decoder for the Greek letters and symbols of ML papers

  Télécharger (PDF)

• Python Language ReferencePython

  Python | Syntax, types, operators, control flow, functions and parameters, LEGB scopes, comprehensions, unpacking, match, f-strings, exceptions (Python 3.12-3.14)

  Télécharger (PDF)

• Python for AI: the Subset That MattersProgramming

  Programming | Types and containers, comprehensions, functions, classes and dataclasses, typing, generators, context managers, async basics, packaging with uv

  Télécharger (PDF)

• Complexity and Big-OAlgorithms

  Algorithms | Counting operations, O / Omega / Theta, time vs space, best / average / worst case, amortised cost, recurrences and the master theorem

  Télécharger (PDF)

• The Data Lifecycle for MLData

  Data folder | Collection, labelling, cleaning, splitting, versioning, monitoring; data-centric AI; datasheets for datasets (Gebru et al.), data cards and Croissant metadata

  Télécharger (PDF)

• Data ContractsData engineering for AI

  Data engineering for AI | What a data contract holds, the Open Data Contract Standard (ODCS v3, as of 2026) annotated, dbt model contracts, breaking vs non-breaking changes, versioning and deprecation windows, enforcement in CI, at write and at read, ownership

  Télécharger (PDF)

• Taxonomy of Machine LearningLearning paradigms

  Learning paradigms | One map of every paradigm along four axes (supervision signal, data access, model type, objective), what each needs, when to use which, and a decision key from problem to paradigm

  Télécharger (PDF)

• Task Types AtlasModel atlas

  Model atlas | Every ML task type as one row: input → output, training loss, typical metric, baseline to beat, where it is taught; an identification key and the pairs people confuse

  Télécharger (PDF)

• What Is Machine Learning? Framing a ProblemMachine learning

  Machine learning | Supervised, unsupervised, self-supervised and reinforcement learning, framing a task, baselines, bias-variance in plain words, no free lunch

  Télécharger (PDF)

• Tabular Feature Pipelines at ScaleTabular and time series

  Tabular and time series 01 | Fit/transform discipline, point-in-time correct training sets (as-of joins), aggregation features on tens of millions of rows with Polars, DuckDB and Spark, memory budgets, train/serve skew and data contracts (feature ideas themselves: data 03; feature-store products: mlops 06)

  Télécharger (PDF)

• Causal Questions and DAGs in PracticeCausal ML

  Causal ML 01 | From a business question to a causal graph: d-separation, the backdoor and frontdoor criteria, minimal adjustment sets, good and bad controls (M-bias, post-treatment variables), testing a DAG against data, DoWhy and dagitty

  Télécharger (PDF)

• Aleatoric vs Epistemic UncertaintyUncertainty quantification

  Uncertainty quantification 01 | Noise in the world vs ignorance of the model: definitions, heteroscedastic regression with a Gaussian negative log-likelihood, the law-of-total-variance split computed on an ensemble, what each kind tells you to do, and where the split breaks down

  Télécharger (PDF)

• Knowledge Representation and LogicSymbolic and probabilistic AI

  Symbolic and probabilistic AI 01 | Propositional and first-order logic, inference rules, CNF and resolution with a worked proof, Horn clauses, semantic networks, frames, description logics and OWL reasoning

  Télécharger (PDF)

• Linear Programming and Duality in PracticeOptimisation and operations research

  Optimisation and OR 01 | Formulating LPs, the graphical method, simplex vs interior point vs first-order (PDLP), the dual as resource prices, complementary slackness, reduced costs, sensitivity ranges, the transportation problem; PuLP, Pyomo, SciPy and HiGHS code

  Télécharger (PDF)

• From Linear Regression to Deep NetworksEvolution of models

  Evolution of models 01 | One dataset of 24 points fitted by mean, linear, polynomial, k-NN, tree, kernel smoother and a small MLP: what each step fixed, what it broke, and bias and variance measured at every step

  Télécharger (PDF)

• Neurons, the Perceptron and the MLPDeep learning

  Deep learning | The artificial neuron, the perceptron and its learning rule, XOR and why depth matters, activation functions, the multilayer perceptron forward pass, parameter counts, universal approximation

  Télécharger (PDF)

• Regression and Classification LossesLosses and training

  Losses and training | MSE, MAE, Huber, log-cosh, quantile (pinball), BCE, cross-entropy with label smoothing, focal and hinge: formulas, gradients, robustness to outliers, class weighting and which one to pick (PyTorch 2.x names)

  Télécharger (PDF)

• Choosing MetricsMetrics and KPIs

  Metrics and KPIs 01 | North star vs guardrail vs diagnostic metrics, what makes a metric good, offline vs online metrics and why they disagree, proxy metrics and how to validate them, KPI driver trees and the metric definition checklist

  Télécharger (PDF)

• The Vision Architecture ZooArchitectures

  Architectures | Reference cards from LeNet to ConvNeXt, ViT, DeiT and Swin, detectors (Faster R-CNN, YOLO, DETR) and segmenters (U-Net, Mask R-CNN, SAM): year, parameters, key idea, what each replaced, receptive fields and U-Net shapes computed

  Télécharger (PDF)

• Graph Data and FeaturesGraph ML

  Graph ML 01 | Modelling data as a graph, storage formats (COO, CSR, adjacency) and their memory, graph types, homophily, node and edge features (degree, clustering, PageRank, centralities, neighbour aggregates) computed by hand, tools from NetworkX to cuGraph, and leakage-safe feature pipelines

  Télécharger (PDF)

• TokenisationLarge language models

  LLM folder | How text becomes token ids: BPE, WordPiece, Unigram and SentencePiece, byte-level BPE, vocabulary size, tokens per language, special tokens, chat templates and the quirks they cause

  Télécharger (PDF)

• Text ClassificationNLP tasks

  NLP tasks 01 | Binary, multi-class, multi-label and hierarchical labelling of text: benchmark datasets, TF-IDF baselines to fine-tuned encoders, SetFit, zero-shot NLI and LLM prompting, label design, thresholds and cascades, confusion analysis, pitfalls in production

  Télécharger (PDF)

• Prompt Patterns LibraryPrompt and eval library

  Prompt and eval library 01 | Twelve copy-paste templates for everyday LLM tasks (brief, few-shot classify, extract, summarise, rewrite, grounded answer, decompose, reason-then-answer, critique-revise, clarify-first, compare, meta-prompt), each with its variables and when to use it; theory in llm 09

  Télécharger (PDF)

• What Is an AI Agent?Agents

  Agents 01 | The agent loop, ReAct, workflows vs agents, autonomy levels, stop conditions, the cost of a loop, when not to build an agent

  Télécharger (PDF)

• AI-Assisted Coding WorkflowsAI for software engineering

  AI for software engineering 01 | Inline, chat, synchronous agent, async agent and headless modes; picking one per task; explore-plan-implement-verify; test-first with agents; task specs; parallel agents in worktrees and the review bottleneck

  Télécharger (PDF)

• Real-Time Voice PipelineVoice and multimodal agents

  Voice and multimodal agents 01 | Cascade (ASR → LLM → TTS) vs half-cascade vs speech-to-speech as an engineering choice, the signal path with echo cancellation, audio formats and token rates, frameworks and realtime APIs as of 2026, prompts written for the ear

  Télécharger (PDF)

• MDPs and Bellman EquationsReinforcement learning

  Reinforcement learning | States, actions, rewards, transitions, return and discount, policies, V and Q functions, Bellman expectation and optimality equations worked by hand, gridworld value iteration preview

  Télécharger (PDF)

• Kinematics and Dynamics EssentialsRobotics in depth

  Robotics in depth 01 | Frames and rotations, homogeneous transforms, forward and inverse kinematics of a 2-link arm worked by hand, the Jacobian and singularities, the manipulator equation, gravity torque, RNEA/ABA/CRBA and Pinocchio / MuJoCo code

  Télécharger (PDF)

• Computer-Vision Tasks, Metrics and ModelsPerception and generation

  Perception and generation | Classification, detection (IoU, NMS, mAP), semantic / instance / panoptic segmentation, keypoints, tracking, OCR; YOLO, DETR, SAM / SAM 2 as of 2026

  Télécharger (PDF)

• Search Engine ArchitectureSearch and recommender systems

  Search and recsys 01 | Crawl and ingest, inverted-index internals (segments, postings compression, refresh and merge, dynamic pruning), sharding and replication, the query-then-fetch serving path and tail latency, Lucene, Elasticsearch, OpenSearch, Solr and Vespa as of 2026

  Télécharger (PDF)

• Physics-Informed Neural Networks (PINNs)Scientific ML

  Scientific ML 01 | The PINN loss (residual, initial/boundary, data terms), collocation and autodiff derivatives, a residual computed by hand, PyTorch code, inverse problems, failure modes and their fixes, variants, and when a classical solver wins (as of 2026)

  Télécharger (PDF)

• Geospatial Data Types and Coordinate Reference SystemsGeospatial and remote-sensing AI

  Geospatial and remote-sensing AI 01 | Raster vs vector, file formats (GeoTIFF, COG, GeoJSON, GeoPackage, GeoParquet, Zarr), datums, projections, EPSG codes, UTM zones, the affine geotransform, distances and areas done right; GeoPandas and rasterio

  Télécharger (PDF)

• Linear and Logistic Regression from ScratchFrom-scratch code

  From-scratch code | Mean-squared-error and log-loss gradient descent in NumPy, checked against least squares, with the learning-rate limit, shape bugs and three extensions (ridge, mini-batch SGD, softmax regression) solved and run

  Télécharger (PDF)

• scikit-learn API ReferenceFrameworks and libraries

  Frameworks and libraries | The estimator / transformer / predictor contract, Pipeline and ColumnTransformer, model selection and search, metrics and scoring strings, common estimators with the hyperparameters to tune, set_output for pandas (scikit-learn 1.x, as of 2026)

  Télécharger (PDF)

• The ML Project LifecycleWorkflows

  Workflows 01 | From business question to monitoring: CRISP-DM, Microsoft TDSP, CRISP-ML(Q), Google's Rules of ML principles; phases, go/no-go gates, deliverables per phase, roles, and a churn project walked through every gate

  Télécharger (PDF)

• The AI Engineer's Toolbox MapTools and platforms

  Tools and platforms | One recommended default and its alternatives for each stage (data, experiment, train, evaluate, deploy, monitor, collaborate), the artefact each stage hands to the next, coverage of the main tools and a stack by team size (opinionated, as of 2026; the full catalogue by task lives in resources 08)

  Télécharger (PDF)

• CI, CD and CT for ML: What ChangesCI/CD for ML

  CI/CD for ML 01 | What differs from software CI/CD: data and model as versioned artefacts, the trigger map (push, tag, schedule, data event, drift alert), the life of a change, pipeline-as-code maturity, delivery metrics for model releases

  Télécharger (PDF)

• ML System DesignMLOps

  MLOps | Requirements, offline vs online, batch vs real-time inference, latency/cost/quality trade-offs, capacity with Little's law, and a full design-interview walk-through (card fraud, feed ranking)

  Télécharger (PDF)

• Event Streams and KafkaStreaming and real-time ML

  Streaming and real-time ML | Topics, partitions, keys and ordering, consumer groups and offsets, retention and compaction, replication and acks, schema registry and compatibility; commands, configs and sizing (Apache Kafka 4.x, KRaft only, as of 2026)

  Télécharger (PDF)

• Export Pipelines: PyTorch to Every RuntimeEdge deployment

  Edge deployment | torch.export and the ONNX exporter, opsets, then Core ML, LiteRT, TensorRT, OpenVINO and ExecuTorch with their commands, operator-coverage and dynamic-shape pitfalls, and a parity check (versions as of 2026)

  Télécharger (PDF)

• GPU Architecture for AICompute infrastructure

  Compute infrastructure | Streaming multiprocessors, warps, tensor cores, HBM and the cache hierarchy, on NVIDIA Hopper as the reference; roofline model and arithmetic intensity to tell memory-bound from compute-bound work

  Télécharger (PDF)

• AI Hardware Types MapAI hardware

  AI hardware | CPU, GPU, TPU, NPU, FPGA, ASIC, neuromorphic, analog, photonic and quantum: how each computes, what it is good and bad at, flexibility vs efficiency, derived metrics from published specs, energy per inference

  Télécharger (PDF)

• Why Networks Matter for AIAI networking

  AI networking | Scale-up vs scale-out, bytes per step for data, sharded, tensor, pipeline, context and expert parallelism, communication/compute overlap and exposed communication, with a 70B tensor-parallel example worked by hand

  Télécharger (PDF)

• The AWS AI Services Map (2026)Cloud

  AWS documentation and announcements to 2026 | AI services, Amazon Q, Bedrock, SageMaker AI, infrastructure; which service for which need; retired and renamed services

  Télécharger (PDF)

• The ML System Design MethodML system design

  System design 01 | Nine steps for interviews and real projects: clarify, requirements, back-of-envelope estimates, data and labels, model, serving, evaluation, monitoring, iteration; a 45-minute time budget and the scoring rubric, on one running example (push-notification ranking)

  Télécharger (PDF)

• AI Safety and AlignmentResponsible AI

  Responsible AI | Outer vs inner alignment, specification gaming and reward hacking, the deceptive-alignment debate, scalable oversight, frontier safety frameworks, dangerous-capability evaluations, AI safety institutes (as of 2026)

  Télécharger (PDF)

• Threat Models for Data Privacy in MLPrivacy-preserving ML

  Privacy-preserving ML | Adversaries by what they see (API, weights, updates, server, operator), attack goals (membership, attribute, reconstruction, linkage), base rates, the DP bound on attacks, LINDDUN per element, and a one-page threat model

  Télécharger (PDF)

• European Union: AI Act, GDPR and Product LiabilityAI law by jurisdiction

  AI law by jurisdiction | AI Act in practice (scope, role changes, enforcement bodies, standards), GDPR for training and automated decisions, the new Product Liability Directive and the EU digital rulebook around AI. Obligations overview as of 2026, not legal advice: check with counsel

  Télécharger (PDF)

• AI Product ManagementAI product and strategy

  AI product and strategy 01 | Discovery for AI features, problem-model fit, PRDs whose acceptance criteria are eval thresholds, deriving the threshold from the cost of errors, the AI product canvas; one running case: AI reply drafts in a helpdesk product

  Télécharger (PDF)

• Designing for Probabilistic OutputsAI UX design

  AI UX design 01 | Output variability in the interface, confidence display options, choosing defaults (suggest vs auto-apply) with an expected-value rule, graceful degradation when the model is slow, unsure or down

  Télécharger (PDF)

• AI in Healthcare and Life SciencesAI by industry

  AI by industry | Care and pharma value chains, use cases with their data, models, clinical KPIs, medical-device rules (EU MDR Rule 11, FDA pathways, PCCP) and the AI Act, build or buy, failure stories, PPV at local prevalence; as of 2026

  Télécharger (PDF)

• Using AI Tools Effectively at WorkAI at work

  AI at work 01 | The tool families as of 2026 and what each is for, task first then tool, five levels of delegation, centaur and cyborg ways of working, the jagged frontier in daily use, a should-I-use-AI decision, session habits and a worked task

  Télécharger (PDF)

• How Open-Source AI Projects WorkOpen source in AI

  Open source in AI 01 | Kinds of open AI artefacts, governance models (BDFL, company-led, foundation, community), roles and how people gain them, repository anatomy, decision channels (issues, RFCs, meetings), funding and health signals; as of 2026

  Télécharger (PDF)

• History of AI, 1943-2011Culture

  AI course, culture | From threshold neurons and the Turing test to Dartmouth, two AI winters, expert systems, backprop, Deep Blue, the SVM decade and ImageNet

  Télécharger (PDF)

• World Models: Video, Physics and PlanningFrontier research 2026

  Frontier research 01 | What a world model must do beyond generating video, the four families as of 2026 (latent control models, interactive video, video foundation models, JEPA), compounding rollout error computed, how to evaluate physics and usefulness, open questions

  Télécharger (PDF)

• AI Certification Map and How to ChooseCertifications

  Certifications | Which AI certification for which role, levels, difficulty, prerequisites, validity and renewal, a decision tree; all exam facts as of 2026, verify on the official page

  Télécharger (PDF)

• Interview Bank: Maths and StatisticsInterview bank

  AI course, interview bank | About 35 interview questions with model answers and what the interviewer listens for: probability, Bayes, hypothesis tests and A/B tests, estimation, linear algebra, calculus and optimisation

  Télécharger (PDF)

• Essential GitHub Repositories for AIResources

  AI course, resources | About 70 repositories by area (frameworks, training, inference, agents, RAG, evaluation, data, education) with owner and what to learn from each; owners checked as of 2026, no star counts

  Télécharger (PDF)

• Glossary - AgentsGlossary

  138 terms collected from the sheets in this collection; [brackets] name the source sheets

  Télécharger (PDF)

Fiches gratuites : Cybersécurité →Fiches gratuites : Cloud →Fiches gratuites : IA →