Data Scientist – Multimodal LLMs (Speech focus)

Full time in AI , in SAAS

Job Detail

  • Job ID 6574
  • Industry  AI

Job Description

📍 Location: Manchester, UK
💰 Salary: Competitive

ConnexAI is launching a new greenfield project to bring speech-to-speech capabilities to large language models. We’re seeking a Data Scientist with a research-driven background in ML, ideally focused on speech or multimodal systems, to help shape this product from the ground up.


What You’ll Do

  • Research and implement cutting-edge methods for integrating audio into multimodal LLMs (ASR, TTS, speech-to-speech).

  • Translate academic research into production-ready models.

  • Train and fine-tune models for performance and scalability.

  • Source and prepare audio/text datasets for training and evaluation.

  • Define evaluation metrics and testing frameworks.

  • Collaborate across research, engineering, and product teams to deploy real-world features.

  • Contribute to improving team processes and fostering innovation.


What You Bring

  • MSc or PhD in ML/Data Science (or equivalent experience).

  • Expertise with LLMs, ASR, TTS, or multimodal model development.

  • Proficiency in Python and PyTorch.

  • Hands-on experience with model training, tuning, and evaluation.

  • Ability to implement ideas from academic papers into practical solutions.

  • Strong communication and collaboration skills.

  • Curious, adaptable, and eager to learn.

Required skills

Other jobs you may like