Senior Platform Engineer, AI Agent Platform & Voice Infra

Rifa AI; United States; Remote; USD 130,000 - 170,000; Full time

Job summary

Rifa AI is hiring an experienced engineer to own the underlying platform that powers AI agents across regulated industries. You will design and ship the runtime, the orchestration engine, and the systems behind Agent Studio, defining agents from procedures and ensuring guardrails.

You’ll also own the voice infrastructure, enterprise security features, and the evaluation framework that validates agent behavior with data-driven experiments.

Qualifications

• 5+ years of hands-on experience building and operating production systems.
• Production experience with cloud platforms and infrastructure as code.
• Real-time systems depth including latency budgets and streaming.
• Experience with observability tooling and incident response.
• Highly autonomous and able to drive outcomes in a high-autonomy environment.

Responsibilities

• Build the agent platform primitives, including runtime, orchestration, and Agent Studio definitions.
• Own the voice infrastructure for real-time audio streaming and low-latency STT/TTS integration.
• Establish enterprise trust through RBAC, audit logs, data isolation, and secure integrations.
• Build the evaluation layer with DeepEval-driven replay, simulations, and AB testing frameworks.
• Develop the observability layer to trace agent decisions and monitor performance.

Skills

Hands-on experience

Kubernetes

Cloud platforms

Real-time systems

Observability tooling

High autonomy

Education

Degree in computer science or related field

Job description

Rifa AI is hiring an experienced engineer to own the underlying platform that powers AI agents across regulated industries. You will design and ship the runtime, the orchestration engine, and the systems behind Agent Studio, defining agents from procedures and ensuring guardrails.

You’ll also own the voice infrastructure, enterprise security features, and the evaluation framework that validates agent behavior with data-driven experiments.