AWS retired the Certified Data Analytics – Specialty exam (DAS-C01) in April 2024 and replaced it with two new credentials. If you've been studying the old exam guide, you've been preparing for the wrong test. This guide covers what the current AWS certification landscape actually looks like for data roles, which credentials employers care about, and the prep courses worth your time.
How the AWS Certification Landscape Changed for Data
Until 2024, the AWS Certified Data Analytics – Specialty was the obvious target for anyone in a data engineering or analytics role. AWS sunsetted it and restructured the data/ML track into:
- AWS Certified AI Practitioner (AIF-C01) — launched September 2024, foundational-level, covers AI/ML concepts and AWS AI services without requiring hands-on engineering depth
- AWS Certified Machine Learning Engineer – Associate (MLA-C01) — launched 2024, replaces the gap between foundational and Specialty, targets practitioners building and deploying ML pipelines
- AWS Certified Machine Learning – Specialty (MLS-C01) — still active, the deepest ML credential AWS offers
- AWS Certified Database – Specialty (DBS-C01) — still active, focused on RDS, DynamoDB, Redshift, Aurora, and database migration
The Solutions Architect – Associate (SAA-C03) remains the most widely recognized AWS certification regardless of role, and many data engineers hold it as their base credential before specializing.
AWS Certification Paths Worth Pursuing in 2026
For Data Engineers
The realistic path if you're working with pipelines, Glue, EMR, Kinesis, or Redshift: start with Solutions Architect – Associate to get comfortable with the broader AWS service surface, then move to Machine Learning Engineer – Associate or Database – Specialty depending on your actual day-to-day work. MLE-Associate is the better choice if you're building ML workflows; DBS-Specialty is better if your role centers on database design and query optimization.
Skipping SAA-C03 and going straight to Specialty-level certs is possible but painful. The Specialty exams assume you can read architecture diagrams and select appropriate services across a wide range, not just the services covered in the specialty domain.
For Data Analysts and BI Engineers
The AI Practitioner (AIF-C01) is a legitimate entry point if your role involves using AWS QuickSight, SageMaker Canvas, or Bedrock but you're not building the underlying infrastructure. It's also the fastest AWS certification to pass — most candidates report 4-6 weeks of prep versus 8-12 weeks for associate-level exams.
If you want something that carries more weight in job applications, go for SAA-C03 instead. Hiring managers across data roles recognize it; AIF-C01 is new enough that it hasn't built that reputation yet.
For Senior Engineers and Architects
AWS Certified Solutions Architect – Professional (SAP-C02) combined with ML Specialty is the stack that shows up most on senior data platform engineer and staff-level job descriptions. These two together signal you can design the infrastructure and build the pipelines — the combination is rarer than either cert alone.
What AWS Certification Exams Actually Test
A common mistake: treating AWS certification prep as memorizing service names. The exams are scenario-based. You'll get a paragraph describing a business requirement and four plausible architectural choices. The wrong answers are usually services that technically work but are more expensive, less scalable, or more operationally complex than the best answer.
For the data-relevant exams specifically:
- SAA-C03: Heavy on S3 storage classes, VPC networking, IAM policies, and when to use RDS vs DynamoDB vs ElastiCache. Data-adjacent but not data-focused.
- ML Engineer – Associate: SageMaker pipelines, feature stores, model deployment patterns, MLflow integration, data preprocessing at scale. Hands-on labs matter more here than for SAA.
- Database – Specialty: Multi-AZ vs read replica trade-offs, DynamoDB partition key design, Redshift distribution styles, Aurora global databases, migration strategies with DMS and SCT.
- ML Specialty (MLS-C01): Statistical depth — you need to know when to use XGBoost vs a neural net, what L1 vs L2 regularization does, and how to handle class imbalance. The most technically demanding AWS certification.
Pass rates are not publicly disclosed by AWS, but practice exam scores are a reliable predictor: if you're consistently scoring above 80% on timed practice exams, you're ready. Below 75%, you're not — take more time.
Top Courses for AWS Certification Prep
AWS Certified Solutions Architect Associate (SAA-C03)
The most thoroughly tested prep course for SAA-C03, with full coverage of all exam domains and hands-on labs that match the question style. Rated 9.6 — unusually high for a technical course, and the rating has held across thousands of reviews. Start here if SAA-C03 is your target.
AWS Certified AI Practitioner Practice Exams (AIF-C01) 2026
Six full-length practice exams for the AIF-C01 credential, updated for 2026 exam objectives. Rated 9.8 — the highest-rated AWS prep resource in this list. The AIF-C01 exam is more conceptual than technical, and timed practice under realistic conditions is the most efficient prep method.
Master PySpark for Data Engineering (AWS, Azure, GCP, Snowflake)
Not a certification prep course — it's a practical skills course covering PySpark on EMR, Glue, and Databricks. Relevant if you're preparing for the ML Engineer – Associate exam and need to actually understand the data processing layer, not just memorize service descriptions. Rated 9.5.
AWS SAP-C02 Practice Exams: 540 Realistic Questions 2026
For candidates targeting the Solutions Architect – Professional exam. SAP-C02 is significantly harder than SAA-C03, and the question volume here (540 questions across multiple timed sets) is enough to expose your weak domains before exam day. Rated 9.5.
Google Cloud IAM and Networking for AWS Professionals
Useful context if you're in a multi-cloud environment or preparing for roles where AWS certification is the baseline but GCP knowledge is also expected. Covers IAM policy models and networking concepts with direct AWS-to-GCP comparisons. Rated 9.7.
Salary and Career Outcomes for AWS-Certified Data Professionals
AWS certification alone doesn't move your salary. What it does is clear the recruiter filter: roles that list "AWS certification preferred" or "AWS certification required" are explicitly screening for the credential. Without it, your application often doesn't make the first cut regardless of experience.
Based on job posting data and salary surveys from 2025-2026:
- Data Engineer with SAA-C03: median US base salary $130,000-$155,000
- Data Engineer with ML Specialty or Database Specialty: $145,000-$180,000
- Solutions Architect – Professional: opens roles at $160,000-$220,000+ at larger companies
- AI Practitioner (AIF-C01): too new to have meaningful salary data; tracks closer to associate-level credentials than Specialty
The ROI calculation is straightforward: exam fees run $150-$300 per attempt, prep courses cost $15-$50 on sale, and preparation typically takes 1-3 months of part-time study. That's a small investment relative to the typical $15,000-$30,000 salary bump when moving into a role that explicitly requires the credential.
One caveat: the salary premium is most pronounced at companies that have standardized on AWS and use certification as a proxy for cloud literacy. At startups or companies using managed services where AWS expertise is less hands-on, the premium is smaller.
FAQ
Which AWS certification should a data engineer get first?
AWS Certified Solutions Architect – Associate (SAA-C03) is the most defensible first choice. It demonstrates broad AWS knowledge, it's recognized across data engineering, DevOps, and platform roles, and it's the expected prerequisite before Specialty-level exams. If you're specifically targeting ML pipelines and already have cloud experience, ML Engineer – Associate is worth considering directly.
Is the AWS Data Analytics Specialty still valid?
AWS retired the DAS-C01 exam in April 2024. If you earned the certification before retirement, it remains on your record and is valid until its expiration date (3 years from issue). It is no longer possible to sit for this exam. The AWS Certified Machine Learning Engineer – Associate and ML Specialty are the closest current equivalents for data-focused roles.
How long does it take to get AWS certified?
For associate-level certifications (SAA-C03, MLE-Associate): most candidates with some cloud exposure take 8-12 weeks studying 10-15 hours per week. Foundational certs like Cloud Practitioner and AI Practitioner take 4-6 weeks. Specialty and Professional exams typically require 12-20 weeks and hands-on AWS experience that's hard to simulate through courses alone.
How much does AWS certification cost?
AWS exam fees: Foundational ($100), Associate ($150), Specialty and Professional ($300). Each failed attempt requires paying the fee again. AWS offers a 50% discount voucher for passing each exam, applicable to a subsequent exam of equal or higher level — so passing SAA-C03 gets you half-off your next attempt at a Professional or Specialty exam.
Do AWS certifications expire?
Yes — all AWS certifications expire after 3 years. Recertification requires passing the current version of the same exam or a higher-level exam in the same path. AWS sends recertification reminders 6 months before expiration. If a cert has been retired (like DAS-C01), you recertify by passing a current equivalent.
Is AWS certification worth it for data roles specifically?
Worth it if: you're targeting companies where AWS is the primary cloud platform, the job description explicitly mentions AWS certification, or you're moving from on-premises data work into cloud data engineering. Less worth it if: your company uses a managed service layer that abstracts AWS (Snowflake, dbt Cloud, Databricks on AWS) or if you're in a pure analytics role where SQL and BI tools are the core skills evaluated.
Bottom Line
The AWS certification landscape for data roles shifted significantly in 2024 with the retirement of the Data Analytics Specialty and the introduction of the AI Practitioner and ML Engineer – Associate credentials. The current best path for most data engineers is SAA-C03 first, then either ML Engineer – Associate or Database Specialty depending on your actual work.
Don't over-index on certification at the expense of hands-on project work. AWS certifications get you past the resume filter; they don't substitute for demonstrated ability to build and ship data pipelines. The candidates who command the highest salaries typically have both — certifications that signal cloud literacy and a portfolio of real work that proves it.
If you're starting prep now, the SAA-C03 course is the highest-leverage starting point for most data roles. If AIF-C01 is your target, the practice exam set is the most efficient path to passing given the exam's conceptual (not hands-on) format.