Software Engineer II, AI

Role overview

Own end-to-end delivery of voice-based AI capabilities used in time-sensitive emergency-response workflows. You will work across model behavior, backend services, APIs, cloud infrastructure, evaluation systems, and the operator-facing web application, with a strong emphasis on safety, reliability, and measurable improvements in conversational performance. The work includes both rapid experimentation and responsible operation of production systems.

Responsibilities

- Take AI-agent capabilities from problem discovery and prototyping through architecture, implementation, testing, deployment, and impact measurement. - Improve latency, turn-taking, interruption handling, speech recognition, multilingual behavior, protocol adherence, and call transfers using production evidence. - Design fail-safe behavior, strengthen resilience and observability, and investigate incidents through root-cause analysis, mitigation, and prevention. - Build automated evaluations, call-replay workflows, production metrics, and controlled traffic experiments to detect regressions and quantify trade-offs. - Contribute across Python services, AI behavior, and React/TypeScript interfaces used to configure and monitor agents. - Use AI-assisted development tools for coding, testing, call analysis, and incident triage while retaining ownership of architecture and safety decisions.

Requirements

- Two to three or more years of backend or full-stack experience delivering production software end to end. - Hands-on experience building AI or LLM applications, with solid distributed-systems, API, and operational-reliability fundamentals. - Strong Python skills, including asynchronous programming, and willingness to work across a TypeScript/React frontend. - Ability to evaluate non-deterministic AI behavior with datasets, automated metrics, qualitative review, controlled experiments, and sound statistical reasoning. - Product-minded judgment: able to simplify ambiguous problems, make evidence-based trade-offs, and own outcomes. - Clear communication, collaborative working habits, and a responsible bias toward action in safety-sensitive systems.

Nice to have

Experience with real-time voice, streaming media, or conversational agents is especially valuable.

Benefits and work setup

The role includes a weekly on-call rotation and travel approximately one or two times per quarter for in-person team gatherings, primarily within Europe. The work environment is flexible and startup-oriented, with competitive compensation, benefits, and equity participation.