Speech & Audio AI Specialist (from int. bpo only)

Beyond Border Consultants; Dadri; On-site; INR 1,400,000 - 2,200,000; Full time

Benefits offered by this job

On-site work

GPU clusters access

Global team exposure

Job summary

Beyond Border Consultants in India seeks a hands-on AI/ASR engineer to design and fine-tune speech recognition and TTS models for noisy call-center environments and multilingual contexts.

You will build voicebot workflows with platforms like Rasa, integrate with telephony APIs and CRM systems, and collaborate with international BPO partners to monitor production pipelines with high uptime.

Qualifications

• Must have hands-on experience with ASR/TTS systems and model fine-tuning.
• Experience building voicebot workflows with telephony or CRM integrations.
• Familiarity with real-world call data and evaluation metrics (WER, F1-score).

Responsibilities

• Design and fine-tune ASR and TTS models for noisy call centers and diverse accents.
• Develop voicebot workflows using Rasa, Dialogflow CX, or similar platforms.
• Train and evaluate speaker diarization, emotion detection, and intent models on real call data.
• Collaborate with international BPO partners to deploy production-grade audio AI pipelines with high uptime.
• Optimize audio preprocessing for low-bandwidth, high-noise channels.
• Conduct A/B testing and report metrics to stakeholders.

Skills

Python

PyTorch

SpeechBrain

Whisper

Google Speech-to-Text

Amazon Transcribe

Librosa

Rasa

Tools

Vosk

Kaldi

Job description

About the Opportunity

A fast-scaling global AI solutions firm operating at the intersection of conversational AI and enterprise contact centers, we build and deploy speech recognition, voicebot, and multimodal audio intelligence systems for Fortune 500 clients. Our team works directly with international BPOs and CX platforms to engineer low-latency, noise-resilient, multilingual voice AI that drives automation, sentiment analysis, and real-time agent assist - powering next-gen customer experience at scale.

Role & Responsibilities

• Design and fine-tune ASR (Automatic Speech Recognition) and TTS (Text-to-Speech) models optimized for noisy call center environments and regional accents.
• Develop voicebot workflows using Rasa, Dialogflow CX, or similar platforms, integrating with telephony APIs and CRM systems.
• Train and evaluate speaker diarization, emotion detection, and intent classification models using real-world BPO call data.
• Collaborate with international BPO partners to deploy and monitor production-grade audio AI pipelines with 99.9% uptime SLAs.
• Optimize audio preprocessing pipelines (noise suppression, VAD, dynamic range compression) for low-bandwidth, high-noise channels.
• Conduct A/B testing of voice AI models and report performance metrics (WER, F1-score, call resolution rate) to product stakeholders.

Skills & Qualifications Must-Have

• Python
• PyTorch
• SpeechBrain
• Whisper
• Google Speech-to-Text
• Amazon Transcribe
• Librosa
• Rasa

Preferred

• Experience with Vosk or Kaldi
• Knowledge of telephony protocols (SIP, RTP)
• Familiarity with BPO operations or CX analytics platforms

Benefits & Culture Highlights

• Work onsite with a global AI team building cutting-edge voice solutions for Fortune 500 clients.
• Access to world-class speech datasets, annotation tools, and GPU clusters for model experimentation.
• Direct exposure to international BPO workflows—ideal for domain specialists aiming to scale voice AI in CX.