Job Description

Voice AI Engineer (Python)

Role Overview

We are looking for a Voice AI Engineer to architect, build, and scale real-time conversational voice AI systems for multilingual users. This role requires deep Python expertise and proven experience building production-grade, ultra-low-latency voice AI applications using LiveKit or Pipecat.

The ideal candidate has hands-on experience with multilingual Indic conversational systems, including Hindi, Marathi, Tamil,Punjabi, Malyalam, etc ... Plus mixed-language/code-switched conversations (e.g., Hinglish/Tanglish), and designing systems that perform reliably with short utterances, interruptions, and natural conversational turn-taking.

Key Responsibilities

• Design, develop, and deploy real-time conversational voice AI systems for production use.
• Build ultra-low-latency voice pipelines integrating ASR, TTS, LLM orchestration, and backend execution.
• Develop Python-based services for audio streaming, session orchestration, voice workflow execution, and backend integrations.
• Build production voice applications using LiveKit or Pipecat.
• Optimize systems for sub-500ms end-to-end conversational latency.
• Implement robust support for:

• multilingual speech interactions
• code-switched conversations
• short utterance detection
• interruptions / barge-in
• silence detection
• turn-taking optimization
• session memory / context handling
• Integrate APIs, CRMs, databases, and internal systems for workflow execution.
• Monitor production systems, debug latency/audio issues, and continuously improve conversation quality.

Mandatory Requirements

• 3+ years of strong Python backend engineering experience
• Proven experience building and shipping production Voice AI / Conversational AI systems
• Hands-on production experience with LiveKit or Pipecat (must-have)
• Demonstrated ability to build ultra-low-latency voice systems with sub-500ms end-to-end response times
• Strong experience with:

• Async Python
• FastAPI
• WebSockets
• event-driven architectures
• real-time audio streaming pipelines
• Hands-on experience integrating:

• ASR providers (Deepgram / Whisper / AssemblyAI / Google Speech / equivalent)
• TTS providers (ElevenLabs / Cartesia / Azure TTS / PlayHT / equivalent)
• LLM orchestration (OpenAI / Anthropic / Gemini / OSS models)
• Proven experience handling:

• Hindi, Marathi, Tamil, and multilingual conversational flows
• code-switched speech (e.g., Hindi-English / Tamil-English mixed speech)
• short utterances ("haan", "bolo", "hmm", "ruk", "ek sec")
• interruption handling / barge-in
• noisy conversational environments
• turn detection optimization
• Strong understanding of:

• conversational state management
• voice agent orchestration
• streaming inference pipelines
• function/tool calling
• latency bottleneck debugging

Preferred Qualifications

• Bachelors degree in Computer Science, Information Technology, Electronics, Artificial Intelligence, Machine Learning, or a related engineering discipline
• Master’s degree in AI / Machine Learning / Computer Science is a plus
• Experience with SIP / WebRTC / telephony integrations
• Experience building inbound/outbound AI calling systems
• Familiarity with Twilio, Daily, Vapi, Retell, or similar infrastructure
• Cloud deployment experience (AWS / GCP)
• Docker / Kubernetes
• Monitoring / observability

More Info

Education