Backend Engineer

About the role:

Clevertap is seeking a Backend Engineer specializing in Workflow Automation using Java and Temporal. You will be responsible for automating internal processes across IT, HR, Finance, Sales, Marketing, and Engineering by building durable workflows.

Key Responsibilities include:

• Building robust workflow systems: Design and implement distributed systems with features like retries, idempotency, long-running state management, and handling in-flight deployments safely using Temporal patterns.
• Leveraging AI Integration: Incorporate LLM capabilities for classification, summarization, intent derivation, and content drafting. Implement human-in-the-loop mechanisms where confidence is low or compliance requires approval.
• System Orchestration: Integrate internal services with third-party SaaS APIs and databases while handling various failure modes effectively.
• Mentorship & Growth: You will learn Temporal development patterns quickly, supported by our code review hooks that guide on versioning and determinism rules. This role offers mentorship in AI tooling integration as well.

What we are looking for:

Must Have Skills:
- 2+ years of professional Java development experience with the ability to discuss production code details.
- Strong understanding of concurrency and error handling principles.
- Working knowledge of distributed systems concepts including retries, idempotency, timeouts, and proper failure mode strategies.
- Experience integrating with external APIs that are not under your direct control.
- Clear communication skills to collaborate effectively with non-engineering stakeholders on process understanding.
Good to have:
- Prior exposure to workflow orchestration tools such as Temporal, Cadence, Airflow, Step Functions, Camunda, or homegrown state machines.
- Hands-on experience calling LLM APIs from backend code.

Measures of Success:

• Migrating manual processes into reliable workflows that teams trust to run autonomously.
• Ensuring system resilience against deploys and crashes without losing data or duplicating work.
  - Implementing observable failures with actionable alerts rather than silent errors.
  - Safely versioning in-flight workflows to avoid orphaned runs or replay issues.
  - Creating debuggable automation tools accessible by the process owners themselves.