AI / Machine Learning Engineer (Voice AI, Agentic Systems & Telephony)
Core
Build and scale the real-time agentic loop connecting automated speech recognition, multi-step LLM function calling, and zero-latency text-to-speech to real-world phone networks for an AI receptionist platform.
Role type
Senior IC machine-learning engineer (voice AI, agentic systems)
Builds
Low-latency conversational AI and robust telephony infrastructure for Swedish and Nordic service businesses
Domain
Telecommunications + Generative AI
Deliverable
production ML models
Required skills
Python, Rust, TypeScript, Agentic AI (tool calling, stateful dialog flow), asynchronous event-driven backends, WebSockets, streaming architectures, speech-to-text, text-to-speech
Preferred skills
SIP, VoIP, FreeSWITCH, Asterisk, local inference engines (vLLM), context caching strategies
Technologies
FreeSWITCH, Twilio, WebRTC, vLLM
Responsibilities
Develop ultra-low-latency agentic loops for voice telephony; Integrate LLM agent workflows with SIP/telephony backends and real-time streaming media pipelines; Benchmark and optimize speech models and frontier LLMs for Swedish conversational flow; Build resilient agent tools for autonomous scheduling and CRM sync; Reduce end-to-end voice roundtrip latency and harden guardrails

