AI Engineer - Voice Agent (Onsite 1pm-10pm)

Full Description

Experience:

2–5 years (hands-on)

Employment Type:

Full-time

Location:

Jail Road, Lahore

About Us

We're on a mission to build intelligent voice agents that can engage in natural, real-time conversations.

These agents are designed to make and receive calls, understand spoken language, and respond with human-like voices using LLMs, advanced STT, and TTS systems.

You'll help shape this future by building scalable, low-latency voice AI systems that interact seamlessly via voice.

What You'll Do

• Voice Agent Development

• Design, implement, and maintain end-to-end voice-agent pipelines: STT → NLU / LLM → TTS.

• Build multi-turn conversational flows, including context management, turn-taking, interruptions, and barge-in.

• Develop prompt-engineering strategies for LLMs to drive dialogues for various call types.

• Integrate voice agent with telephony systems (SIP, Twilio, WebRTC) to support real voice calling.

• AI / ML Engineering

• Train, fine-tune, and serve models for intent detection, NLU, or dialog management.

• Work with LLMs (OpenAI, Llama, etc.) for generating responses in voice conversations.

• Optimize models for inference cost and latency (quantization, batching, efficient serving).

• Implement MLOps practices: version control, model deployment, monitoring, and retraining pipelines.

• Speech & Voice Tech

• Use high-quality STT (e.g., Whisper, Deepgram, Google) and TTS (e.g., ElevenLabs, Azure, Google) systems.

• Implement voice activity detection (VAD) and manage audio streaming / chunking for real-time voice interaction.

• Ensure voice output is natural, expressive, and appropriate for conversation scenarios.

• System Design & Infrastructure

• Design scalable, low-latency architecture for real-time voice interactions.

• Deploy voice agents in cloud environments (e.g., AWS) and set up serving infrastructure.

• Build monitoring, logging, and analytics around conversation quality, latency, and errors.

• Ensure data privacy and voice data security.

What We're Looking For

Must-Have

• 2+ years of hands-on experience in AI / ML, especially conversational or voice AI systems.

• Demonstrated experience building voice agents (STT + TTS + dialog management).

• Strong Python experience and deep learning / ML framework knowledge.

• Experience with LLMs (prompting, fine-tuning) or NLU models.

• Knowledge of telephony integrations (SIP, Twilio, or WebRTC).

• Familiarity with MLOps practices and deploying ML models to production.

• Ability to build low-latency, scalable systems.

Nice-to-Have

• Previous projects or deployed systems as a calling / voice agent.

For instance, job descriptions for similar roles include working with STT, TTS, LLMs, and telephony. Jobaaj+1

• Experience with large-scale voice / conversational AI engines (Rasa, LangChain, etc.).

• Knowledge of emotion detection, voice biometrics, or multi-speaker voice modeling.

• Familiarity with deployment on cloud voice / telecom infrastructure.

• Understanding of AI ethics, data privacy (especially voice), and compliance for voice data.

Why Join Us?

• You'll

own the voice agent architecture

, building from the ground up.

• Work on

state-of-the-art AI+voice systems

combining STT, LLMs, and TTS.

• High impact

: your work will define how human-like, useful, and scalable our voice agents are.

• Great learning opportunity in MLOps, low-latency inference, and real-time dialogue systems.

• Close-knit team where your technical decisions will directly shape the product.

How to Apply

Please submit:

Your

resume / CV

A summary of your

past voice-AI or conversational agent projects

— especially deployments or real-time voice systems

(Optional but recommended) Links or demo video of any voice agent / conversational AI work you've done