AI Engineer – Voice, LLM & Real-Time Systems

🚀 We’re hiring: AI Engineer - Voice, LLM & Real-Time Systems

📍 Belgrade (Hybrid) | 🇨🇭 Swiss client (Coorpix)

For our partner Coorpix, a Switzerland-based company building cutting-edge AI solutions, we are looking for an AI Voice Engineer to join a new team in Belgrade.

👉 This is a unique opportunity to work on a production AI platform in partnership with Cisco, currently being implemented in Swiss courts – building real-world voice AI systems from the ground up.

Your Responsibilities

• Design and build real-time voice agents that autonomously handle phone conversations
• Develop low-latency streaming pipelines (STT → LLM → TTS)
• Integrate telephony systems via Azure Communication Services (ACS) (SIP, RTP, session handling)
• Implement call logic (routing, timeouts, voicemail, busy signals)
• Optimize latency, responsiveness, and overall performance
• Build systems tailored for Swiss German dialect recognition, including phonetic matching and noisy 8kHz audio environments
• Implement natural conversation mechanics (turn-taking, interruption handling, barge-in)
• Design scalable systems handling high concurrency (Redis, distributed architectures)
• Build observability systems (monitoring, tracing, incident metrics)
• Extend LangGraph-based orchestration and structured conversation flows
• Ensure strict PII protection with enforced scrubbing pipelines (no raw sensitive data exposure to LLMs)
• Apply security and compliance best practices for voice data
• Collaborate closely with AI and infrastructure teams to deliver end-to-end systems

Your Profile

• Strong engineering skills in Python, with experience in FastAPI, asyncio, and real-time systems
• Proven experience with low-latency architectures and streaming pipelines
• Hands-on with STT/TTS providers (Azure Speech, ElevenLabs, or similar)
• Familiarity with telephony systems (ACS, SIP, RTP, WebRTC)
• Experience building distributed, scalable systems
• Background in LLM-based systems, orchestration (LangGraph/LangChain), and structured prompts
• Solid understanding of performance and reliability engineering
• Knowledge of audio processing (sample rates, codecs, telephony constraints) is a plus
• Ability to own systems end-to-end
• Fluent in English

What we offer

• Hybrid work from Belgrade office two times per week (Located at Belgrade Waterfront)
• Work on production AI systems deployed in real-world environments
• Direct impact on the development of next-generation voice AI systems
• High ownership and technical responsibility
• A culture that values autonomy, clarity and impact
• Flexible workload (60–100%) and hybrid work setup (Switzerland / EU)
• Long-term growth opportunities as our AI platform evolves and scales.
• A workplace where technology meets purpose and your work leaves a visible mark