Audio AI Developer

Deevia Software India; Bengaluru; On-site; INR 1,200,000 - 1,700,000; Full time

Job summary

Deevia Software India seeks a highly motivated AI Audio/Speech Developer with strengths in Deep Learning, Speech Recognition, and Large Language Models. The role involves designing, training, fine-tuning, and deploying AI models, implementing research papers, and crafting state-of-the-art speech and language solutions.

Ideal candidates will independently design and optimize models from scratch, work on LLM/SLM architectures, and collaborate with teams to integrate AI into production.

Qualifications

• 2–5 years experience in speech-to-text and language model development.
• Hands-on experience with ASR, NLP, and transformer-based architectures.
• Proficiency in Python and ML frameworks (PyTorch, TensorFlow).

Responsibilities

• Design and develop Speech-to-Text (ASR) and language processing solutions using Deep Learning techniques.
• Build, train, fine-tune, and optimize transformer-based models from scratch.
• Implement and reproduce advanced research papers related to speech recognition, NLP, and Generative AI.
• Work on LLM and SLM architectures, model adaptation, quantization, and deployment optimization.
• Develop and improve speech recognition pipelines for accuracy, latency, and resource efficiency.
• Perform model evaluation, benchmarking, and performance tuning.
• Collaborate with cross-functional teams to integrate AI models into production applications.
• Stay updated with the latest advancements in Speech AI, LLMs, and Generative AI technologies.

Skills

Deep Learning

Transformer Architecture

LLMs

ASR

Speech-to-Text

Python

PyTorch

TensorFlow

Model optimization

Quantization

Fine-tuning

Tools

PyTorch

TensorFlow

Job description

Experience

2 to 5 years of experience, with at least 2+ years of hands-on experience in Speech-to-Text (ASR) and Language Model development.

Job Summary

We are looking for a highly motivated AI Audio/Speech Developer with strong expertise in Deep Learning, Speech Recognition, and Large Language Models. The ideal candidate should be capable of independently designing, training, fine-tuning, and optimizing AI models from scratch, implementing research papers, and developing state-of-the-art speech and language solutions.

Key Responsibilities

• Design and develop Speech-to-Text (ASR) and language processing solutions using Deep Learning techniques.
• Build, train, fine-tune, and optimize transformer-based models from scratch.
• Implement and reproduce advanced research papers related to speech recognition, NLP, and Generative AI.
• Work on LLM and SLM architectures, model adaptation, quantization, and deployment optimization.
• Develop and improve speech recognition pipelines for accuracy, latency, and resource efficiency.
• Perform model evaluation, benchmarking, and performance tuning.
• Collaborate with cross-functional teams to integrate AI models into production applications.
• Stay updated with the latest advancements in Speech AI, LLMs, and Generative AI technologies.

Required Skills

Technical Skills

• Strong knowledge of Deep Learning, Machine Learning, and Neural Network architectures.
• Deep understanding of Transformer Architecture and Large Language Models (LLMs).
• Hands-on experience with:
  • Automatic Speech Recognition (ASR)
  • Speech-to-Text systems
  • Language Models
  • Transformer-based architectures
• Experience in training, fine-tuning, and optimization of ASR models.
• Hands-on experience with:
  • LLM fine-tuning
  • Quantization
  • Model optimization
  • Parameter-efficient tuning techniques
• Strong programming skills in Python.
• Experience with PyTorch and/or TensorFlow.
• Experience implementing research papers and converting them into production-ready solutions.