Note: The job is a remote job and is open to candidates in USA. ICON INTERNATIONAL GROUP is seeking an Airflow Support Engineer focused on Security & Operations. This role is primarily a support and operational position requiring excellent troubleshooting skills and documentation discipline to address client needs within SLA requirements.
Responsibilities
- Provide support for Apache Airflow in production environments
- Troubleshoot Airflow failures, including restart loops, scheduler lag, task queue issues, executor problems, and dependency conflicts
- Manage AWS production experience, specifically ECS on EC2 for containerized Airflow
- Implement security hardening measures such as RBAC, secrets backends, and TLS
- Conduct CVE triage and vulnerability justification, documenting findings for compliance review
- Utilize Python for reading and debugging DAGs and providers
- Understand Docker/containerization fundamentals
- Communicate effectively with clients while adhering to SLA-bound response and resolution requirements
Skills
- 4+ years hands-on with Apache Airflow in production (not just “used it in a pipeline”) — scheduler, executors, DAG lifecycle, backfills
- Strong troubleshooting of Airflow failures: restart loops, scheduler lag, task queue issues, executor problems, dependency conflicts
- AWS production experience, specifically ECS on EC2 (containerized Airflow) — sizing, deployment, networking
- Security hardening: RBAC, secrets backends, TLS, and reducing Airflow's provider surface / CVE footprint — this is FedPoint's core pain, so weight it heavily
- Comfortable with CVE triage and vulnerability justification — assessing exploitability, documenting findings for compliance review
- Python proficiency (Airflow is Python-native; needed to read/debug DAGs and providers)
- Docker/containerization fundamentals
- Strong written communication — client-facing support with SLA-bound response/resolution
- Experience supporting federal / government contractor environments; familiarity with FISMA/NIST 800-53/800-171 concepts (support context, not as a certifier)
- Kubernetes (some clients run Airflow there even if FedPoint is ECS)
- CI/CD, infrastructure-as-code (Terraform/CloudFormation)
- Prior SLA-based support or managed-services experience (vs. pure engineering) — knows what “2-hour response” discipline actually means
- Monitoring: CloudWatch, Prometheus, Grafana
Company Overview