Staff Engineer – Real-Time Voice AI & Simulation (SF)

Cekura; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Benefits offered by this job

Top-of-market comp

Meaningful equity

US visa sponsorship

Daily meals

Relocation support

Job summary

Cekura in San Francisco is building the core simulation engines, evaluation systems, and real-time voice infrastructure for enterprise-grade agent reliability. This is a hands-on, full-stack role at a small, senior team.

You’ll own end-to-end features from design to production, collaborating with founders, and shipping thousands of conversations with controlled personas and noisy channels. The role demands long hours in person, fast pace, and a strong bias for action, with competitive pay,

Qualifications

• Strong coding ability in Python, TypeScript, or Go.
• Experience with distributed systems, real-time infrastructure, or LLM-based products.
• Strong written and verbal communication.

Responsibilities

• Build the testing and simulation engine to run thousands of conversations across voice, chat, and phone.
• Push the frontier of agent evaluation with LLM-powered evaluators and adversarial testing.
• Develop applied audio and speech research to separate noise from speech and extract paralinguistic signals.
• Own real-time voice infrastructure including SIP, WebRTC, STT/TTS pipelines and providers.
• Ship end-to-end features from design to production with full-stack ownership.
• Shape architecture and coding standards in a small senior team.

Skills

Python

TypeScript

Go

Distributed systems

Real-time infrastructure

LLM evals

Reliability engineering

Tools

Twilio

SIP

WebRTC

Vapi

Retell

LiveKit

Pipecat

STT/TTS pipelines

Job description

Cekura in San Francisco is building the core simulation engines, evaluation systems, and real-time voice infrastructure for enterprise-grade agent reliability. This is a hands-on, full-stack role at a small, senior team.

You’ll own end-to-end features from design to production, collaborating with founders, and shipping thousands of conversations with controlled personas and noisy channels. The role demands long hours in person, fast pace, and a strong bias for action, with competitive pay,