Cloud-related work can involve infrastructure, software, security, data, support or design. Choosing a direction is easier when you identify the responsibility behind the title. A list of platform names or certificates cannot answer that question alone.
These six directions are a research map, not a ranked demand list or a promise that one training route qualifies you for every role.
1. Infrastructure and platform operations
Investigate how the team provisions, changes and maintains its environment. Ask about review, access boundaries and operational support. A personal lab should use your own authorised resources and clear cost controls.
2. Cloud-focused software development
Explore how applications are built and maintained for the team’s environment. Show a manageable project and explain configuration, testing and failure handling without claiming that a demonstration is production ready.
3. Reliability and automation
Look at repeatable changes, monitoring and the work needed to make services dependable. Ask what on-call duties and escalation support apply. The title may overlap with platform, DevOps or SRE responsibilities.
4. Cloud security work
Choose a specific responsibility such as secure configuration, access governance or security engineering. Practise only in authorised environments. Do not treat familiarity with a checklist as proof of broad security expertise.
5. Cloud data work
Investigate how data is moved, prepared and made available reliably. Ask about quality checks, ownership and operational responsibilities instead of assuming the role is only about writing transformations.
6. Architecture and customer-facing solutions
Explore the experience needed to understand requirements and explain trade-offs. These roles may expect substantial technical and domain judgement; do not assume an “architect” title is an entry-level destination after a short course.
Choose the next practical step
Choose a direction and compare actual vacancies by tasks, level and support. The DevOps and SRE guide offers questions for one common overlap. Keep learning claims specific and use safe demonstrations rather than publishing real credentials or experimenting on an employer's systems without permission.
