AWS Certification Roadmap: Every Path Explained

Here’s a number that stopped me mid-scroll last month: over 1.42 million AWS certifications are now active worldwide, held by more than 1.05 million individuals, according to AWS’s own certification page. That’s a global professional language, not a niche credential. So if you’ve been staring at the AWS Certification Roadmap wondering which of the twelve exams deserves your evenings and weekends, you’re asking the right question. I’ve sat six of these exams over the past four years, failed one badly, and rebuilt my study approach from scratch. This guide is the AWS certification path I wish someone had handed me on day one.

Why Does the AWS Certification Roadmap Matter More in 2027?

Cloud spending hasn’t slowed down, and neither has the confusion over which certificate to chase first. Amazon Web Services still commands roughly 30 to 32 percent of the global cloud infrastructure market, which means AWS skills remain the default language of cloud engineering teams from Berlin to Singapore.

AWS has also reshaped its catalogue. It added the Generative AI Developer – Professional exam, renamed SysOps Administrator to CloudOps Engineer – Associate, and retired the Machine Learning – Specialty on March 31, 2026. A well-planned certification path now has to account for AI exams too, which is why so many candidates feel overwhelmed before booking a single exam.

AWS Certifications List 2027: All 12 Exams at a Glance

AWS Certification Levels: Building Your Certification Path Step by Step

AWS groups its certifications into four levels: Foundational, Associate, Professional, and Specialty. Understanding this ladder is the real starting point of any AWS certification roadmap. The foundational tier (Cloud Practitioner or AI Practitioner) is optional for anyone already in tech but useful for people moving into cloud-adjacent roles from sales, finance, or project management.

From there, most candidates move to an associate exam matched to their target job: Solutions Architect Associate for generalists; Developer Associate for coders; CloudOps Engineer Associate for operations folks; and Data Engineer Associate for anyone leaning toward data engineering.

The professional tier is demanding. Solutions Architect Professional and DevOps Engineer Professionals both assume hands-on exposure to production environments, not just flashcards. Professional exams have no formal prerequisites, but SAP-C02 and DOP-C02 assume deep operational experience with systems already running at scale.

Passing multiple-choice questions is possible without that experience; retaining it under real production pressure is not.

Five Career Tracks Inside the AWS Certification Path

Rather than listing exams in isolation, it helps to see how this certification path maps onto actual job families. Here’s how I’d group the tracks I get asked about most often:

1. Cloud Infrastructure and Architecture

• Core path: Cloud Practitioner → Solutions Architect Associate → Solutions Architect Professional
• Best for engineers who provision, scale, and cost-optimise workloads across accounts
• This is the highest-volume path in job postings, and its architecture patterns transfer across cloud providers
• Typical study time: 4 to 8 weeks per Associate exam, 10 to 16 weeks for the Professional tier

2. Cloud Security

• Core path: Solutions Architect Associate → Security Specialty
• Best for engineers hardening identity, network boundaries, and encryption
• Employers increasingly pair this track with multi-cloud experience, since breach exposure rarely stays inside one provider
• Typical study time: 6 to 10 weeks after the associate foundation is in place

3. Data Engineering and Machine Learning

• Core path: Data Engineer Associate → Machine Learning Engineer Associate
• Best for professionals building ingestion pipelines, feature stores, and production ML systems
• The Machine Learning Engineer Associate replaces the retired ML Specialty as the main ML credential
• Typical study time: 6 to 12 weeks per exam, longer if you’re new to data pipelines

4. Generative AI

• Core path: AI Practitioner → Machine Learning Engineer Associate → Generative AI Developer Professional
• Best for developers building applications on foundation models, Amazon Bedrock, and AI agents
• This is the fastest-growing branch of the AWS certification roadmap as generative AI workloads move to production
• Typical study time: 3 to 4 weeks for the AI Practitioner, 10 to 14 weeks for the Professional exam

5. Site Reliability Engineering and Operations

• Core path: CloudOps Engineer Associate → DevOps Engineer Professional
• Best for engineers running on-call rotations, automating recovery, and keeping systems available under load
• Typical study time: 8 to 14 weeks, factoring in hands-on lab practice

How Much Does AWS Certification Cost?

Every pass earns a 50 percent discount voucher for your next exam, so plan your sequence with the voucher in mind. Over a full certification path, that saves several hundred dollars. Check current local pricing on the AWS Certification page before booking.

AWS Certification Salary: What Each Level Pays in 2027

Salary is usually the real reason people commit to an AWS certification roadmap in the first place, especially in data engineering and machine learning engineering, where pay has climbed fastest. These figures are pulled from multiple 2027 sources rather than a single survey, since salary reporting varies by region.

Numbers compiled from Skillsoft’s IT Skills and Salary survey and other 2027 salary reports. One pattern holds across every source: certified professionals report a 20 to 30 percent salary increase over non-certified peers, an effect that sharpens at the Professional and Specialty tiers.

Which AWS Certification Should I Get First?

I get some version of “where do I start” almost every week, so here’s the honest breakdown I give people:

• No IT background at all: start with Cloud Practitioner. It’s not required, but it builds vocabulary you’ll need later and earns a 50 percent voucher toward your next exam.
• Interested in AI but non-technical: AI Practitioner is the better first badge.
• Already working in tech: skip Cloud Practitioner and go straight to Solutions Architect Associate. It carries far more weight with hiring managers and still covers the fundamentals.
• You write production code: Developer Associate maps more directly to your daily work.
• You’re aiming at operations or reliability work: CloudOps Engineer Associate, then DevOps Engineer Professional once you’ve logged real incident-response hours.
• You want security or compliance work: Solutions Architect Associate for the architecture foundation, then Security Specialty.
• You’re drawn to pipelines and models: Data Engineer Associate builds the base for data engineering, while Machine Learning Engineer Associate is the more direct route into ML.

Certifications are a job-search tool, not a collection hobby. Stacking five credentials before your first cloud role usually wastes months better spent on hands-on projects.

Multi-Cloud Strategy and Where AWS Fits

No certification path exists in isolation anymore. Most mid-size and large organizations now run some form of multi-cloud strategy, whether by design or acquisition history, and hiring managers value candidates who understand how AWS services interoperate with other providers. That doesn’t mean parallel certifications from every vendor; a documented cross-provider architecture matters more than a stack of badges.

Treat AWS depth as the anchor skill, then layer cross-platform networking and identity concepts on top as you move into senior roles.

What Did I Learn Sitting for Six of These Exams?

A personal note, since it matters more than another bullet list: my first attempt was the old Machine Learning Specialty, and I failed because I studied services instead of scenarios. AWS exams rarely ask, “What does this service do?” They ask, “Which services solve this oddly-worded business problem?”

Once I switched to reading whitepapers and rebuilding small systems in a sandbox account instead of memorising service lists, my record went from one failure to five straight passes, including Solutions Architect Professional on the first try. If there’s one thing I’d tell anyone starting their own AWS certification roadmap today, it’s this: build something small and break it on purpose.

Common Mistakes That Derail an AWS Certification Path

A few patterns show up again and again among people I’ve mentored through this process:

• Chasing every specialty exam before landing a first cloud job. It burns budget without payoff. A Security Specialty adds little if your target is data engineering.
• Ignoring the employer’s stack. Azure shops won’t reward AWS badges the way an AWS shop will.
• Attempting Professional exams without real operational scars. This is the top cause of first-attempt failure.
• Studying for a retired exam. Check that your course covers the current code (SOA-C03, not SOA-C02; MLA-C01, not the ML Specialty).
• Treating recertification as an afterthought. Every AWS credential expires after three years, and passing a higher exam automatically renews the lower ones beneath it.

Conclusion: Your Next Move on the AWS Certification Roadmap

The AWS Certification Roadmap isn’t about collecting twelve badges. It’s about choosing two or three that compound toward a role you want, whether that’s cloud architecture, security, data engineering, generative AI or site reliability engineering.

Start with the level matching your current experience, build real projects alongside your study plan, and broaden into multi-cloud as you gain seniority. The exams reward people who’ve broken things and fixed them, not people who memorised the fastest.