AI/ML Engineer

AI / ML Engineer – Multilingual Voice AI Systems

Location: Noida- Onsite
Experience: 3–6 YearsType: Full-Time/ContractualReporting To: Solution Architect / Technical Director

We’re building a multilingual, AI-driven voice assistant for phone-based booking via telephony APIs e.g. (Ozonetel / Exotel / Twilio / Knowlarity). The system must interact naturally in multiple global languages, making it a first-of-its-kind solution in our product stack.

Role Overview

We're seeking a practical AI/ML Engineer to lead the speech AI and multilingual NLP components of a voice automation platform. You’ll design and integrate Speech-to-Text, LLM/NLU, and Text-to-Speech systems that support dynamic phone conversations in multiple languages.

Key Responsibilities

• Design, train, or integrate multilingual NLP pipelines to support dynamic conversations using:
• LLMs (OpenAI GPT, LLaMA 3, Ollama, or Azure OpenAI)
• STT/TTS services (Google Cloud, Azure Speech, Whisper, etc.)
• Configure language detection, intent recognition, and fallback flows
• Create multilingual prompt engineering logic and fine-tune models (if needed)
• Ensure STT/LLM/TTS works effectively with real-time voice streams via telephony APIs e.g. (Ozonetel / Exotel / Twilio / Knowlarity)
• Support integration of fallback IVR-style flows when AI fails
• Work closely with .NET engineers and backend team to wire up API-based workflows
• Assist in evaluating trade-offs between cloud APIs vs self-hosted models
• Optimize voice latency, transcription quality, and conversational UX
• Help set up language-specific voice personas with TTS (tone, pitch, accent)

Must-Have Skills

• Experience building or integrating conversational AI or voicebots
• Strong knowledge of NLP, multilingual text handling, and LLM usage
• Hands-on integration with:
• OpenAI / Hugging Face / Ollama / LangChain
• Google Speech / Azure STT & TTS / Whisper
• Good understanding of international language encoding, accents, and noise handling
• Python fluency (NLTK, transformers, speech SDKs, etc.)
• Ability to work with APIs (REST, JSON, async jobs)

Good-to-Have Skills

• Knowledge of telephony APIs (Ozonetel / Exotel / Twilio / Knowlarity)
• Familiarity with audio pipelines (e.g., audio capture, normalization, latency tuning)
• Experience with language-specific nuances like:
• Arabic TTS tokenization
• Hindi/Urdu word splitting
• Nigerian Pidgin vs standard English handling
• Prior experience with voice UX design or multilingual chatbot building
• Familiarity with Whisper fine-tuning or custom STT model training

Qualifications

• Bachelor’s / Master’s in CS, Data Science, AI, or equivalent
• 3–6 years in AI/ML product development with focus on NLP, STT, or TTS
• Past experience in multilingual bot or voice system is highly preferred

Job Types: Full-time, Contractual / Temporary, Freelance
Contract length: 3 months

Work Location: In person