ServiceNow
Staff Voice AI Engineer – SRE/DevOps
$176K–$308K
Santa Clara, CALIFORNIA, United States · Hybrid · Staff · Full-time · Posted Oct 6, 2026
- Experience
- 4+ years operating cloud-native workloads and 4+ years of development experience; 2+ years building and operating Voice AI applications is a strong asset.
- Visa sponsorship
- All employment is contingent upon obtaining any export license or other approval that may be required by relevant export control authorities.
About the job
Build and operate cloud-native infrastructure and software for Voice AI platforms, covering reliability, deployment, observability, and LLM integration.
What you'll do
- Develop operational solutions for Voice AI, including deployment, monitoring, high availability, disaster recovery, and continuous integration.
- Own software from design through implementation, test automation, and delivery, applying engineering best practices.
- Mentor colleagues and promote knowledge-sharing across telecom and AI engineering.
What they ask for
- 4+ years operating highly available distributed cloud-native workloads on container platforms.
- 4+ years of development experience with Python, GoLang, Java, or similar languages.
- Experience integrating applications on top of LLMs (Prompt Engineering Agents and Skills).
- Experience building distributed systems with cloud-native software; experience with software-defined networking, infrastructure as code, and configuration management.
Nice to have
- 2+ years building and operating Voice AI applications is a strong asset.
- Experience with DevOps tooling (e.g. Helm / Ansible / Kubernetes / Prometheus /Splunk/ GitLab CI) is considered an asset.
Pay
$176,100 - $308,200 USD base salary
Tools
Kubernetes, Python, GoLang, Java, Helm, Ansible, Prometheus, Splunk
AI focus
Integrate LLMs into voice platforms and real-time communication systems, build Voice AI applications, and apply AI tools and integrations to work processes, decision-making, and problem-solving.
Model-written summary · the original posting is authoritative