MLOps / Infrastructure Engineer 

About eFiche

eFiche is a Rwanda-based digital health company building AI-powered clinical tools that support nurses and midwives at the frontline of primary health care. Our work spans ambient clinical documentation, AI-driven decision support, and risk flagging — all designed to reduce cognitive load on health workers and improve the quality of care delivered to patients across Rwanda's health system. We operate at the intersection of AI engineering, clinical practice, and health system strengthening, and we are scaling fast.

About the Role

As our AI platform grows to serve an expanding network of health facilities nationwide, we are looking for an experienced MLOps / Infrastructure Engineer to own the reliability, security, and scalability of the infrastructure that underpins everything we build.

The MLOps / Infrastructure Engineer owns the infrastructure of eFiche's AI services layer — from development and staging environments through to production — ensuring the AI platform is reliable, secure, and performant as we expand to national coverage.

Key Responsibilities

• Design, build, and maintain automated build, testing, and deployment pipelines for eFiche's AI services
• Manage cloud and on-premise infrastructure, including server provisioning, containerisation, and environment management
• Own AI model serving infrastructure — deployment, versioning, latency monitoring, and rollback protocols
• Ensure infrastructure compliance with Rwanda's NCSA cybersecurity standards and the project's data governance framework
• Monitor system uptime, performance, and incident response; maintain ≥95% uptime SLA
• Support security audits and penetration test remediations
• Ensure the AI services layer remains available and performant when communicating with national health system API endpoints

Required Profile

• 5+ years of experience in MLOps, infrastructure engineering, or DevOps roles

Position Details

• Reporting to: Technical Lead (AI/EMR)
• Location: Kigali, Rwanda
• Type: Full-time