UnleashX is building the AI workforce. Companies use our platform to deploy digital teammates that talk, act, and deliver outcomes across sales, support, and operations. These are not assistants. They are accountable teammates measured on performance and results.
💼 Role Details
Title : AI Engineer (Voice AI)
Location : Gurugram
Engagement : Full-time
Compensation : ₹24–36 LPA + ESOP
🧠 What You’ll Do
- Build and improve real-time voice agents for sales, support, and operations use cases
- Work with ASR, TTS, LLMs, and dialog orchestration frameworks to enhance voice quality and response accuracy
- Integrate multiple providers (Deepgram, Sarvam, ElevenLabs, OpenAI) into a single modular voice pipeline
- Reduce latency, improve interrupt handling, and make conversations more natural
- Own agent evaluation : call audits, response quality scoring, accuracy improvements
- Work closely with product and customers to understand real call behaviours and fix friction points
- Contribute to building our voice infra : WebRTC, streaming sockets, and call routing flows
⚙️ What You Bring
3–6 years experience as AI / ML / LLM engineer working with voice or conversational systemsHands-on with LLM prompting, finetuning, embeddings, and RAGSolid understanding of STT and TTS models and how to tune them for natural speechProficient in Python and NodeExperience working with real-time audio streaming (WebRTC, WebSockets) is a strong advantageAbility to debug live calls and improve agent behaviour fastOwnership mindset and willingness to iterate directly with users🏆 What Success Looks Like
Voice agents handle interruptions smoothly and respond naturallyCall success rates and customer completion rates improve month over monthToken, latency, and infra costs reduce through better engineering and model choiceCustomers start trusting agents as teammates, not toolsIf you are someone who :
✅ Builds fast and learns faster
✅ Cares about product experience, not just model performance
✅ Wants to create the future of how businesses talk to customers
Then you’ll do great here.