The short answer: choose the work, then choose the exam
The best AWS certification is not automatically the most advanced or the most popular. It is the credential whose candidate profile matches the decisions you already make—or the decisions you are prepared to practise repeatedly. A project manager who needs to discuss AI risk has a different starting point from a developer building event-driven applications. A data engineer operating pipelines needs different evidence from a solutions architect designing multi-account networks. A security specialist should not choose a foundational exam merely because it appears first in a diagram.
Start by writing one sentence: “I want to be trusted to…” Complete it with a real responsibility. Examples include “explain cloud value to customers,” “design secure and resilient applications,” “deploy and troubleshoot workloads,” “build governed data pipelines,” or “operate production generative-AI systems.” That sentence is a better filter than salary lists, social-media popularity or a promise that one certificate is best for everyone.
The current official AWS exam-guide directory contains 12 certifications: two Foundational, five Associate, three Professional and two Specialty paths. AWS describes Foundational credentials as broad understanding, Associate credentials as role-specific technical validation, Professional credentials as complex advanced work, and Specialty credentials as deep technical expertise in a particular domain. There is no requirement to collect every level. A useful path is selective: it closes a skills gap, produces practical evidence and supports a role you genuinely want.
Do not begin with “Which badge is easiest?” Begin with “Which capability must I demonstrate?” Use the official candidate description, build the capability, then use an exam as an external validation point.
A five-question decision framework
Use these questions before purchasing training or scheduling an exam. They prevent the two most common mistakes: choosing a level that is too shallow for your experience, or choosing an advanced title before you have the practical foundation to learn from it.
1. What work do you want to perform?
Choose one primary direction: business cloud literacy, architecture, development, operations, data engineering, machine learning, generative AI, DevOps, networking or security. If you select five directions at once, you are not choosing a certification—you are describing a long-term learning program. Pick the next responsibility first.
2. What evidence do you already have?
List projects, incidents, migrations, pipelines, applications, designs or customer decisions you can explain. Evidence can come from employment, labs or a carefully documented portfolio, but you should know what happened, why the design was chosen and how you verified the result. A badge without explainable evidence is fragile in an interview.
3. How broad or deep is the role?
Foundational exams build shared language. Associate exams validate a defined technical role. Professional exams combine domains and organizational constraints. Specialty exams go deep into a domain. Broad architecture and deep security are both valuable, but they require different preparation.
4. What is your realistic practice environment?
A certification plan must fit the time, account access, budget and support you can sustain. Prefer small, repeatable labs with budgets and cleanup steps. If your chosen exam expects production-style reasoning, reading alone will not create the necessary judgment.
5. What will you do after the exam?
Decide whether the next action is a job portfolio, a project assignment, a higher-level credential or deeper work in the same role. The exam date should be a checkpoint in a skills plan, not the end of learning. AWS services, exam guides and operational practices continue to change.
Use a readiness gap, not a calendar guess
Compare your evidence with the official domain tasks. Mark each task as “can explain,” “can perform with help,” or “can perform and troubleshoot.” Your study duration follows from that gap. Two candidates targeting the same exam can reasonably need very different preparation time.
Understand the four certification levels
Foundational: vocabulary, value and core decisions
Foundational does not mean worthless or effortless. It means the exam is designed around broad understanding rather than a specific implementation role. CLF-C02 covers cloud concepts, security and compliance, technology and services, and billing, pricing and support. AIF-C01 focuses on AI and ML fundamentals, generative AI, foundation-model applications, responsible AI, security and governance. These exams help non-specialists participate in informed conversations and give beginners a structured map.
A foundational certificate is useful when it removes confusion that blocks collaboration. It is less useful when an experienced engineer collects it instead of developing the role-specific skills an employer expects. If you can already explain the foundational scope and have relevant hands-on experience, assess an Associate guide directly.
Associate: role-specific technical capability
Associate paths are where the roadmap branches by work: architecture, development, operations, data engineering and machine learning engineering. They share cloud foundations but test different decisions. For example, an architect compares designs across requirements; a developer reasons about application behavior; an operations engineer uses evidence to deploy and restore; a data engineer handles pipelines and governance; an ML engineer owns data-to-model workflows.
Do not treat Associate certifications as interchangeable. Choose the role whose labs you are willing to build and whose troubleshooting you want to practise. SAA-C03 is broad, but it is not a universal prerequisite for every other Associate exam.
Professional: complex systems and accumulated judgment
Professional exams assume that individual service facts are already familiar. The difficulty comes from combined constraints, long scenarios and tradeoffs across teams, accounts, reliability, security, operations and cost. SAP-C02 emphasizes advanced architecture; DOP-C02 emphasizes automated delivery and operations; AIP-C01 emphasizes production generative-AI implementation. Choosing Professional because of the title, without relevant practice, usually leads to memorisation instead of durable skill.
Specialty: deep domain expertise
ANS-C01 and SCS-C03 validate advanced networking and security expertise. A Specialty exam is appropriate when the domain is part of your real responsibility and you can diagnose it, not simply recognize terms. A solutions architect may later specialize, while an experienced network or security practitioner may approach the Specialty path through a different sequence.
All 12 current AWS certifications, explained
The profiles below summarize who each path serves, when to choose it, what credible evidence looks like and how to prepare. They are planning guidance, not substitutes for the official exam guides. AWS can revise names, codes, domains, services, languages and policies, so re-check the linked official directory before booking.
AWS Certified Cloud Practitioner
- Who it fits
- Learners who need a broad understanding of cloud value, AWS services, security responsibilities, billing, pricing and support without being assessed as an architect or developer.
- When to choose it
- Choose CLF-C02 when you are new to cloud, work in sales, recruitment, finance, project coordination or customer-facing delivery, or need a common AWS vocabulary before moving into a technical role. It is also a sensible first checkpoint for a technical beginner who has not yet built or operated workloads.
- Evidence to build
- Your evidence should be more than memorised service names. You should be able to explain why an organization uses elasticity, which party secures what under shared responsibility, how Regions and Availability Zones support resilience, and how common compute, storage, database, networking and cost tools fit a simple workload.
- Preparation direction
- Study the four official domains, build a small budget-controlled AWS environment, compare services by use case, and practise explaining decisions in plain language. Move to SAA-C03, DVA-C02, SOA-C03 or a role-specific Associate certification when you can connect foundational concepts to hands-on work.
AWS Certified AI Practitioner
- Who it fits
- People who work with AI initiatives and decisions but do not necessarily build production models: business analysts, product and project managers, sales professionals, marketers, support teams and technology leaders.
- When to choose it
- Choose AIF-C01 when your work requires credible understanding of artificial intelligence, machine learning, generative AI, foundation-model use cases, responsible AI, security and governance. It is not a substitute for the implementation depth expected from an ML engineer or generative-AI developer.
- Evidence to build
- You should be able to distinguish predictive ML from generative AI, select a suitable business use case, recognize prompt and foundation-model concepts, discuss responsible-AI risks, and identify why data protection, monitoring and human review matter. The official guide also expects familiarity with AWS AI services and core cloud services used around them.
- Preparation direction
- Combine concept learning with small demonstrations in Amazon Bedrock or related AWS learning environments. Practise describing business value, limitations, evaluation, security and cost rather than only naming features. Candidates who later implement solutions can progress toward MLA-C01, DEA-C01 or AIP-C01 after building the necessary engineering experience.
AWS Certified Solutions Architect – Associate
- Who it fits
- Cloud engineers, solution designers, developers, administrators and technical consultants who make architecture choices across security, resilience, performance and cost.
- When to choose it
- Choose SAA-C03 when your work involves translating requirements into AWS designs. It is often the strongest first Associate target for someone who needs a broad technical view because it connects identity, networking, compute, storage, databases, integration, disaster recovery and cost optimization.
- Evidence to build
- You should be able to justify a design rather than repeat service definitions. For example, explain when private subnets need a NAT gateway or VPC endpoint, when RDS Multi-AZ solves availability rather than read scaling, how IAM roles remove embedded credentials, and how an architecture changes when recovery time, recovery point, latency or cost requirements change.
- Preparation direction
- Use the official four-domain guide as your scope. Build a multi-tier application, configure identity and network boundaries, test failure and recovery, inspect monitoring data, and compare at least two valid designs for each requirement. ITCertPath connects the SAA-C03 course, detailed decision guides and original timed practice from this roadmap.
AWS Certified Developer – Associate
- Who it fits
- Software developers and cloud application engineers who build, deploy, secure, observe and troubleshoot applications using AWS services and development tooling.
- When to choose it
- Choose DVA-C02 when your daily work is application delivery rather than broad infrastructure administration. The path fits developers using AWS SDKs, event-driven services, serverless compute, application configuration, CI/CD, observability and identity controls inside cloud applications.
- Evidence to build
- A credible candidate can implement authentication and authorization, handle retries and idempotency, choose synchronous or asynchronous integration, manage application secrets and configuration, deploy safely, and use logs, traces and metrics to diagnose failures. Knowing an API name is not enough; you need to understand runtime behavior and failure modes.
- Preparation direction
- Build several small applications rather than one large demo. Include a Lambda and API workflow, a queue-driven process, a data store, secure configuration, automated deployment and observability. Read the current DVA-C02 guide before booking because the assessed services and emphasis evolve as AWS development practices change.
AWS Certified CloudOps Engineer – Associate
- Who it fits
- Cloud operations engineers, systems administrators, platform teams and support engineers responsible for deploying, managing, monitoring, automating and troubleshooting AWS workloads.
- When to choose it
- Choose SOA-C03 when your strongest evidence comes from keeping systems reliable and controlled. It fits people who work with monitoring, incident response, patching, backup, networking, identity, governance, automation, deployment and cost-aware operations rather than primarily writing application features.
- Evidence to build
- You should be able to use operational evidence to choose an action: distinguish a metric alarm from a log query, diagnose connectivity, restore protected data, automate repeatable changes, apply permissions safely, and evaluate the operational effect of scaling or configuration decisions. The role requires judgment during both normal operations and incidents.
- Preparation direction
- Operate a realistic environment for several weeks. Create dashboards and alerts, simulate a failed instance or blocked network path, restore from backup, automate configuration, inspect audit activity and document a runbook. Treat troubleshooting as a sequence of evidence checks rather than guessing a service change.
AWS Certified Data Engineer – Associate
- Who it fits
- Data engineers and analytics platform practitioners who ingest, transform, orchestrate, store, govern, monitor and optimize data workloads on AWS.
- When to choose it
- Choose DEA-C01 when you design or operate data pipelines and data stores. It is a better fit than a general architecture exam when your work centers on batch and streaming ingestion, transformation, orchestration, data quality, cataloging, access controls and the cost or performance of analytical systems.
- Evidence to build
- Your portfolio should show an end-to-end pipeline: how data arrives, where it lands, how schema and quality are handled, how transformations are orchestrated, how consumers discover and query data, and how failures are retried or recovered. Be ready to explain partitioning, file formats, encryption, least privilege, monitoring and cost decisions.
- Preparation direction
- Build one batch pipeline and one streaming or near-real-time pipeline. Include malformed records, late data, schema change, backfill, access separation and observability. Compare serverless and provisioned choices. Use the official DEA-C01 domains to identify missing skills instead of studying every AWS analytics service equally.
AWS Certified Machine Learning Engineer – Associate
- Who it fits
- ML engineers, data scientists moving into production, data engineers and software engineers who prepare data, train models, deploy inference and operate ML workflows on AWS.
- When to choose it
- Choose MLA-C01 when you implement and maintain ML solutions rather than only discuss AI concepts. AWS describes a target candidate with practical SageMaker and related AWS experience, plus experience in an adjacent engineering or data role. AIF-C01 can establish vocabulary, but it does not replace this hands-on foundation.
- Evidence to build
- You should be able to prepare trustworthy data, perform feature engineering, select and refine modeling approaches, deploy endpoints or batch inference, orchestrate pipelines, monitor model and data behavior, secure resources and control cost. Your evidence should include why a model or deployment choice fits the requirement and how you detect degradation.
- Preparation direction
- Create a reproducible ML workflow from raw data to monitored deployment. Track experiments, manage artifacts, separate training and inference permissions, automate a pipeline and define a response to drift or performance failure. Review the current four-domain MLA-C01 guide and its in-scope services before final revision.
AWS Certified Solutions Architect – Professional
- Who it fits
- Experienced architects and senior cloud engineers who design complex multi-account, hybrid, migration, modernization and governance solutions across organizational boundaries.
- When to choose it
- Choose SAP-C02 after you have repeatedly made and defended architecture decisions in real environments. It is not simply a harder SAA-C03. Professional scenarios combine organizational complexity, migration constraints, security boundaries, reliability, performance, operations and financial tradeoffs in the same decision.
- Evidence to build
- Your experience should include multi-account strategy, identity federation, network connectivity, workload migration, disaster recovery, governance at scale, service quotas, operational ownership and cost optimization. You should recognize when an apparently technical answer creates excessive operational work or conflicts with a business constraint.
- Preparation direction
- Study through architecture reviews. For each scenario, state requirements, constraints, assumptions, failure modes, alternatives and the reason for the selected design. Revisit Well-Architected principles, Organizations and control patterns, hybrid networking, migration approaches and multi-Region design. Use long scenarios to practise eliminating answers that solve only one part of the problem.
AWS Certified DevOps Engineer – Professional
- Who it fits
- Experienced DevOps, platform, site-reliability and cloud operations engineers who automate delivery and operations across distributed AWS environments.
- When to choose it
- Choose DOP-C02 when you own delivery systems and operational automation at scale. Candidates should be comfortable with infrastructure as code, deployment strategies, observability, incident response, policy enforcement, configuration management, reliability improvement and secure multi-account operations.
- Evidence to build
- Strong evidence includes reusable deployment pipelines, automated testing and rollback, traceable configuration, centralized logging and monitoring, event-driven remediation, recovery exercises and measurable reliability outcomes. You must understand why a control belongs in a pipeline, account boundary, runtime or operational process.
- Preparation direction
- Build and break a delivery platform in a safe environment. Implement blue/green or canary delivery, create an alarm that triggers action, test rollback, rotate a secret, recover a workload and review audit evidence. Map every exercise to current DOP-C02 domains and document operational tradeoffs rather than memorising service combinations.
AWS Certified Generative AI Developer – Professional
- Who it fits
- Experienced developers and AI engineers who integrate foundation models into production applications and own safety, security, evaluation, operations and cost outcomes.
- When to choose it
- Choose AIP-C01 when you have moved beyond demos and prompt experiments. AWS positions this exam for candidates with substantial cloud or application experience and hands-on experience implementing generative-AI solutions. It is designed around production delivery, not general awareness of AI terminology.
- Evidence to build
- You should be able to design retrieval-augmented generation, manage data and embeddings, integrate models and agents, evaluate quality, apply guardrails, protect sensitive information, observe production behavior, troubleshoot failures and optimize latency and cost. Your evidence must show controlled deployment and measurable evaluation, not only a working chatbot.
- Preparation direction
- Build a small production-style application with authenticated access, governed data ingestion, retrieval, model evaluation, safety controls, tracing and cost visibility. Test adversarial and failure cases. Review the official AIP-C01 domains because generative-AI services and recommended patterns change quickly.
AWS Certified Advanced Networking – Specialty
- Who it fits
- Network architects and senior cloud engineers who design and troubleshoot complex AWS, hybrid and multi-account network architectures at scale.
- When to choose it
- Choose ANS-C01 when networking is already a professional strength. The exam expects deeper reasoning about routing, connectivity, segmentation, DNS, traffic engineering, security, automation and troubleshooting than an Associate architecture path. It should not be selected only because Specialty sounds more advanced.
- Evidence to build
- You should be able to interpret routes and packet paths, choose between connectivity patterns, design resilient hybrid links, separate accounts and environments, control ingress and egress, diagnose asymmetric or overlapping networks, and explain how DNS and application delivery interact with network architecture.
- Preparation direction
- Create diagrams and packet-flow narratives for every lab. Build multi-VPC connectivity, test private service access, analyze routing and security controls, inspect flow evidence and simulate a failure. The most useful preparation asks where a packet starts, every decision it crosses, and why it succeeds or fails.
AWS Certified Security – Specialty
- Who it fits
- Security engineers, cloud security architects and experienced practitioners who protect workloads, identities, data and governance processes on AWS.
- When to choose it
- Choose SCS-C03 when security is your primary responsibility and you already understand AWS architecture. The path covers more than enabling security services: it requires identity design, data protection, infrastructure security, detection, incident response, governance and the ability to investigate evidence.
- Evidence to build
- A strong candidate can trace effective permissions, protect encryption keys, design logging and alerting, separate preventive and detective controls, contain an incident, preserve evidence and choose remediation that does not create another exposure. You should be comfortable with both technical controls and operational ownership.
- Preparation direction
- Build a security lab around identities, encryption, network boundaries, centralized logs, findings and response. Deliberately create safe misconfigurations, detect them, investigate the evidence and remediate them. Use the current SCS-C03 guide as the boundary and connect every service choice to a threat, control objective and response process.
Role-based paths: practical sequences, not mandatory ladders
AWS does not require you to complete a fixed sequence before most exams. The paths below are learning patterns, not prerequisites. Skip a step when you can already demonstrate its outcomes. Add a step when your practical gap is larger than the exam title suggests.
Cloud beginner or non-technical stakeholder
CLF-C02 → role-specific Associate. Build cloud language, security responsibility, service categories and cost awareness. Then choose architecture, development, operations, data or ML based on the work you want. A business professional focused specifically on AI may use AIF-C01 instead of, or after, CLF-C02.
Solutions architect
SAA-C03 → SAP-C02. Build broad architecture judgment first, then add organizational complexity, migration, multi-account governance and advanced resilience. Add ANS-C01 or SCS-C03 only when networking or security becomes a deep responsibility.
Application developer
DVA-C02 → DOP-C02 or AIP-C01. Begin with secure cloud application delivery. Move toward DevOps when you own delivery platforms and operations, or toward generative AI when you implement governed production AI applications. SAA-C03 is useful when design responsibility becomes broader.
Cloud operations or platform engineer
SOA-C03 → DOP-C02. Develop monitoring, troubleshooting, deployment, recovery and automation skills. Progress when you design reusable delivery and operations systems across teams or accounts. Add architecture or security depth where the platform role requires it.
Data engineer
DEA-C01 → architecture, ML or AI specialization. Start with governed pipelines and data stores. Add SAA-C03 for broader system design, MLA-C01 for model operations or AIP-C01 for production generative-AI systems. Choose the branch from actual project responsibilities.
Machine-learning or AI practitioner
AIF-C01 for awareness; MLA-C01 for ML engineering; AIP-C01 for production GenAI. These credentials do not form a compulsory ladder. AIF-C01 suits concept and business decisions. MLA-C01 suits model workflows. AIP-C01 suits experienced developers integrating foundation models into production.
Network engineer
SAA-C03 or SOA-C03 foundations → ANS-C01. If you are new to AWS, first learn how services, accounts and workloads use networks. Experienced cloud network professionals can assess ANS-C01 directly but should still check gaps in AWS-specific routing, hybrid connectivity and service networking.
Security engineer
SAA-C03 or operational foundations → SCS-C03. Learn how AWS workloads are built before securing them at depth. Experienced cloud security practitioners may go directly to SCS-C03 after confirming architecture, identity, data protection, infrastructure security, detection and incident-response readiness.
A preparation system that produces skills, not only scores
A good study plan has four connected loops: scope, learn, build and review. It should expose weak reasoning early and produce evidence you can revisit after the exam. The following system adapts to any AWS certification.
Freeze the target exam and current guide
Record the exact exam name, code and date you reviewed the guide. Save the official domain outline and candidate description. Do not mix materials from retired versions or assume a course title proves current alignment. Check announcements again before scheduling.
Run a diagnostic without chasing the score
Use an original practice set or official question set to reveal unfamiliar domains. For every wrong answer, label the cause: missing concept, misread requirement, weak service comparison, no hands-on experience or poor elimination. A score alone does not tell you what to change.
Build a domain evidence map
Create a table with each official task statement, an explanation in your own words, one practical exercise and one failure or tradeoff. A green task is one you can explain and perform. A yellow task needs support. A red task has no evidence. Spend time according to the map, not equally across chapters.
Learn concepts through decisions
Convert facts into comparisons. Ask when to choose the service, when not to choose it, what fails, what it costs operationally, how it is secured and how it is observed. Scenario questions reward the best fit among plausible options, so your notes should preserve tradeoffs.
Build small labs with cleanup and evidence
Set a budget, use least privilege, record architecture and remove resources after each exercise. Capture configuration, test results and a short explanation. A lab is complete when you can state the requirement, design, verification and lesson—not when a console wizard shows success.
Practise under timing, then review slowly
Timed practice helps pacing, but learning happens during review. Explain why the correct option fits every requirement and why each distractor fails. Return to the official task and a hands-on exercise when the same reasoning gap repeats.
Use a readiness gate before booking
Require stable performance across different questions, explainable decisions across every domain and successful hands-on exercises. Do not book because you memorised one bank. The real exam can differ in wording, question types, coverage and raw scoring from independent practice.
A realistic weekly rhythm
Use short weekday sessions for concepts and retrieval practice, one longer session for a lab, and one review session for mixed scenarios. For example: Monday maps one domain; Tuesday compares services; Wednesday builds; Thursday troubleshoots; Friday explains decisions aloud; Saturday completes mixed practice; Sunday reviews errors and plans the next gap. The exact number of weeks depends on your starting evidence, not a marketing promise.
Keep an error log with four fields: requirement missed, mistaken assumption, correct decision rule and action taken. If you repeatedly choose a highly available design when the question asks for disaster recovery, your gap is not the name of a service—it is distinguishing availability from recovery objectives. That insight should change the next lab.
What hands-on evidence should look like
Hands-on preparation does not require a giant production clone. It requires controlled experiments that make important behavior visible. The scenarios below can be scaled to your experience and target role.
Resilient web workload
Define availability and recovery requirements, draw a multi-AZ design, secure network paths, select data protection, create monitoring and test what happens when one component fails. Compare a lower-cost alternative and explain what risk it accepts.
Event-driven order flow
Build an API that validates a request, publishes work asynchronously and stores a result. Add idempotency, retries, dead-letter handling, structured logs and least-privilege access. Trigger one failure deliberately and trace it through the system.
Observable and recoverable service
Create meaningful metrics, logs, alarms and a runbook. Simulate resource exhaustion or blocked connectivity, diagnose from evidence, restore service and record time to detection and recovery. Automate one repeated response safely.
Governed pipeline
Ingest a small dataset, handle malformed records, partition output, catalog it, restrict access and monitor freshness. Perform a backfill and schema change. Explain how the design controls cost, quality and recovery.
Evaluated model workflow
Prepare data, track an experiment, deploy or run inference, capture quality and operational metrics, and define a rollback or retraining response. For generative AI, add evaluation criteria, guardrails, protected data handling and cost observation.
Trace access and packet flow
Start with a denied identity or network request. Trace trust, permissions, routes, stateful and stateless controls, DNS and encryption. Fix the smallest necessary control, verify the result and record why a broader permission would be unsafe.
How to use mock tests responsibly
A mock test is a diagnostic instrument, not a source of live exam content. Good practice measures decisions across the public scope, provides explanations and makes weak domains visible. It should not claim to reproduce the official exam or guarantee a pass. Avoid websites that market copied questions, certainty or “latest dumps” as a shortcut; memorisation can leave you unable to perform the job and may expose you to inaccurate or unauthorized material.
Use the first attempt before you feel ready. Its purpose is to establish a baseline. Review every answer, including correct guesses. Study and build the weakest domain, then take a different set. A repeated high score on the same questions measures memory. Improvement across new scenarios and practical work is stronger evidence.
ITCertPath mock tests use original learning scenarios. They are independent practice, not official AWS exam questions. For current scope and policies, always use AWS documentation as the authority.
AWS exam voucher and booking checks
Do not purchase a voucher until the certification decision is stable. AWS states that certification exams are scheduled through an AWS Certification Account and delivered by Pearson VUE. AWS also directs exam-voucher purchases to Pearson as an authorized third-party distributor. A training provider may help you understand options, but the exact voucher terms, testing region and redemption process must be verified before payment.
| Check | Why it matters | Evidence to keep |
|---|---|---|
| Exact exam and code | A similar certification name or retired version may not match your booking. | Current official exam page and written quote. |
| Testing country and currency | Availability, taxes, local pricing and redemption can vary by region. | Country-specific scheduling screen and final total. |
| Expiration date | AWS warns that voucher expiration dates cannot be extended simply because an appointment changes. | Voucher terms and an exam date comfortably before expiry. |
| Assignment and transfer rules | Team vouchers may have management or assignment conditions. | Issuer documentation naming the purchaser and candidate process. |
| Retake or campaign terms | Promotions have dates, eligibility and same-exam conditions; they are not permanent benefits. | Official campaign page captured at purchase. |
| Cancellation and rescheduling | AWS policies include deadlines and limits. Changes close to the appointment can forfeit fees. | Confirmation email and current policy page. |
AWS currently describes online proctoring and test-center delivery through Pearson VUE. Before an online appointment, test the computer, network, webcam, microphone and room requirements using the official instructions. Ensure the name in your Certification Account matches acceptable identification. For a test center, verify location, arrival time and ID requirements in the confirmation.
AWS also states that certified individuals can access a 50% exam-discount benefit for recertification or another AWS exam. Treat this as an account benefit with its own terms, not as a public coupon. Certification validity is currently three years, and recertification requirements should be checked in the official policy before planning a renewal path.
Turn the roadmap into your next action
If you are undecided, send ITCertPath your current role, years of relevant experience, target role, hands-on AWS exposure, preferred timeline and country of testing. Good guidance may recommend that you delay booking, change the target or build a project first. It should not promise a pass or pressure you into a voucher before the exam is confirmed.
For SAA-C03 candidates, the ITCertPath learning cluster already connects a structured course, architecture decision guides, original full-length mock tests and AWS voucher checks. Other certification paths can use this roadmap and the AWS hub while focused resources are expanded. The daily content program will deepen one intent-driven cluster at a time rather than publishing hundreds of near-duplicate pages.
Target role · current experience · certification considered · practical AWS work completed · strongest and weakest domains · preferred test country · whether you need training, practice, voucher guidance or all three.
Frequently asked questions
Which AWS certification is best for a complete beginner?
CLF-C02 is the broad cloud-foundation option. AIF-C01 is the foundational choice for people focused on AI concepts and business use cases. A technical beginner who already builds systems may instead start with an Associate path after checking the official candidate profile and completing hands-on preparation.
Should I take CLF-C02 before SAA-C03?
AWS does not require CLF-C02 as a prerequisite for SAA-C03. Take it first when you need cloud vocabulary, shared-responsibility, billing and service-category foundations. If you already work with AWS and can explain those topics, use the SAA-C03 guide to assess whether direct Associate preparation is appropriate.
Is AWS certification enough to get a cloud job?
No credential guarantees a job. Certification can validate knowledge, but employers also evaluate practical evidence, troubleshooting, communication and relevant experience. Pair preparation with labs, a documented project, architecture explanations and examples of decisions you can defend.
How do I choose between SAA-C03, DVA-C02 and SOA-C03?
Choose by daily responsibility. SAA-C03 emphasizes architecture decisions, DVA-C02 emphasizes cloud application development, and SOA-C03 emphasizes deployment, operations and troubleshooting. Review each official candidate description and compare it with the work you already do or want to practise.
Which AWS certification should a data engineer choose?
DEA-C01 is the direct data-engineering path. It covers pipelines, data stores, operations, governance and performance. SAA-C03 can add broad architecture knowledge, while MLA-C01 fits data professionals who move into production machine-learning workflows.
Which AWS certification covers generative AI?
AIF-C01 validates foundational AI and generative-AI knowledge for people who may not build solutions. AIP-C01 is a Professional certification for experienced practitioners implementing production generative-AI applications. Choose based on implementation experience, not only interest in AI.
Can I buy an AWS exam voucher for another person?
AWS directs exam-voucher purchasers to Pearson, its authorized third-party distributor, and describes vouchers as a way to fund team certification. Always verify the testing country, exact exam, validity, assignment rules and refund or extension terms before payment. A voucher expiration date cannot be extended merely because an appointment is changed.
How long is AWS certification valid?
AWS states that its certifications are valid for three years from the date earned. Candidates must meet the current recertification requirement before expiration. Check the official recertification policy because available paths and higher-level renewal relationships can change.
Are ITCertPath mock tests official AWS exam questions?
No. ITCertPath practice uses original learning scenarios and is not the official AWS exam. Use it to identify weak concepts and practise decisions, then use the current AWS exam guide and official AWS preparation resources as the scope authority.
Can any website guarantee that I will pass or rank first in search?
No responsible training provider can guarantee an exam result, and no SEO provider can guarantee a first-position Google ranking. Useful preparation combines current official scope, practical learning, honest practice and consistent review. Search visibility depends on relevance, quality, technical accessibility, competition and user response over time.
Official AWS sources used for this guide
These primary sources were checked on 30 August 2026. AWS controls its certification names, scope, policies, prices and delivery options. Re-check the official pages before making a purchase or booking decision.