About the Company: Devnagri is a AI company dedicated to personalizing business communication and making it hyper-local to attract non-English speakers. We address the significant gap in internet content availability for most of the worlds population who do not speak English. For more detail - Visit www.devnagri.com

What Youll Do:

Develop real-time voice bots with WebRTC, LiveKit, WebSockets, and telephony integration.

Build ASR pipelines with Whisper, VAD, turn detection, and inference systems.

Fine-tune and deploy Generative AI models using Ollama, vLLM, HuggingFace.

Optimize TTS streaming using Orpheus, Spark, Fish Audio or similar open-source models.

Implement Python-based inference microservices and ML Ops pipelines for scalable deployment.

Ensure low-latency, high-performance, and secure streaming voice solutions.

You Should Have:





3 years RELEVANT experience in real-time AI/voice streaming development mostly on open source models WITHOUT using openAI, Azure or Amazon Lex.

Solid Python skills and experience with ASR, TTS, WebRTC, LiveKit, WebSockets.

Knowledge of Generative AI, model fine-tuning, quantization, and HuggingFace hosting.

Experience with MLOps, scalable microservices, and low-latency streaming systems.

Familiarity with telephony integration and real-time conversational AI.

Why Join Us:

Work with cutting-edge AI tech, open-source voice models, and a passionate team solving real-world communication challenges.

Interview Process:

Screening Round then Shortlisting >> 1-2 technical round >> Managerial round >> HR Closure

with your short success story in in building real-time streaming AI voice bots.

Must have skills

STT TTS Rag model Self hosting

📌 AI Voicebot Developer (Uttar Pradesh)
🏢 Caprus IT
📍 Uttar Pradesh