Key Responsibilities
Conduct applied research to develop, train, and optimize speech recognition models incorporating NLP, deep learning, and generative AI techniques.
Design and experiment with LLMs and RAG models for improved conversational AI systems.
Drive finetuning and domain adaptation strategies for enhancing model performance on multilingual and low-resource speech data.
Work with agentic AI frameworks to enable advanced workflow orchestration and stateful agent interactions.
Collaborate with cross-functional teams, including other AI researchers and scientists, to seamlessly integrate features and functionalities.
Collaborate with engineering teams to implement and deploy scalable cloud-based AI pipelines and services.
Analyse and preprocess large-scale speech datasets, ensuring high-quality data management for model training.
Publish research findings in internal and / or external conferences,



actively participate in hackathons and contribute to the scientific community.
Mentor junior researchers and contribute to knowledge sharing and technical leadership within the team

Required Skills
Expertise in speech and audio signal processing, in domains such as speech coding, speech enhancement, automatic speech recognition (ASR), or text-to-speech (TTS) and/or neural synthesis
Experience with deep learned representations and self-supervised learning
Experience with capture, transmission, and rendering of speech and audio
Experience developing algorithms and systems for mobile voice telephony and/or voice interfaces
Embedded software development for ARM, DSP, and/or neural signal processing hardware
B.Tech/M.Tech/Ph.D. Computer Engineering
Robust understanding of hardware-aware optimization (CPU, GPU, edge ac

📌 Speech Scientist Kolkata
🏢 Mihup Communications
📍 Kolkata