Senior AI Production Support Engineer

Ema; Plymouth; On-site; GBP 85,000 - 110,000; Full time

Job summary

Ema Unlimited Inc. is seeking a Senior Support Engineer to own the health, reliability, and trustworthiness of deployed agentic AI systems in production.

This role sits at the intersection of AI behavior, workflow orchestration, enterprise integrations, and customer trust. You will partner with Engineering, Product and Implementation to prevent incidents, diagnose complex Sev-1/2 issues, and provide actionable workarounds while engineering fixes are implemented.

Qualifications

• 6–9+ years in senior technical support, reliability, or production engineering roles.
• Experience supporting enterprise SaaS platforms in production.
• Strong understanding of APIs and integrations.
• Knowledge of distributed systems, logs, traces, and diagnostics.
• Experience handling escalations and incidents with customers.
• Customer-facing experience with senior stakeholders.

Responsibilities

• Own customer-reported issues from intake to resolution, meeting SLAs.
• Debug agentic workflows end-to-end and trace failures across agents, tools, and humans-in-the-loop.
• Communicate clearly with engineers and customers during incidents; provide realistic recovery timelines.
• Write high-signal incident reports and feedback to engineering/product.
• Improve runbooks, playbooks, and diagnostics; mentor junior SREs.

Skills

Debugging instincts

Root-cause analysis

Customer communication

Incident management

Cross-functional collaboration

Tools

APIs & integrations

Distributed systems

Logging & tracing

Cloud knowledge (AWS/GCP/Azure)

Job description

Ema Unlimited Inc. is seeking a Senior Support Engineer to own the health, reliability, and trustworthiness of deployed agentic AI systems in production.

This role sits at the intersection of AI behavior, workflow orchestration, enterprise integrations, and customer trust. You will partner with Engineering, Product and Implementation to prevent incidents, diagnose complex Sev-1/2 issues, and provide actionable workarounds while engineering fixes are implemented.