Real-Time Speech ML Engineer - ASR/TTS Pipelines

ConnexAI; Manchester; On-site; GBP 70,000 - 110,000; Full time

Job summary

ConnexAI in Manchester is seeking a Machine Learning Engineer to maximise the efficiency of ASR and TTS components within our Agentic AI pipeline. You will collaborate with scientists and engineers to bridge research and production, delivering scalable inference systems.

The role focuses on building low-latency ML pipelines, deploying models, and ensuring reliability across cloud platforms using Docker, Kubernetes, and modern ML frameworks. Strong PyTorch and Python skills are essential.

Qualifications

• MSc degree in a related field with 2+ years hands-on experience in Speech or Generative AI.
• Strong command of Python and PyTorch in production environments.
• Experience deploying ML models to cloud platforms with containerised workflows.

Responsibilities

• Liaise with ASR/TTS leads and engineers to deploy models on Hopper and Blackwell GPU architectures.
• Maintain biweekly progress in a sprint-based research environment.
• Build low-latency real-time inference pipelines for speech recognition and synthesis.
• Optimise model performance across cloud platforms using TensorRT and ONNX.
• Develop robust API services using Python frameworks (e.g., FastAPI).
• Manage containerisation and orchestration with Docker and Kubernetes.
• Ensure observability with Prometheus, Grafana, and OpenTelemetry.
• Document technical work and research findings.

Skills

Python

PyTorch

Generative AI

Communication skills

Education

MSc in CS/ML or related field

Tools

Docker

Kubernetes

Triton Inference Server

ONNX

TensorRT

FastAPI

Job description

ConnexAI in Manchester is seeking a Machine Learning Engineer to maximise the efficiency of ASR and TTS components within our Agentic AI pipeline. You will collaborate with scientists and engineers to bridge research and production, delivering scalable inference systems.

The role focuses on building low-latency ML pipelines, deploying models, and ensuring reliability across cloud platforms using Docker, Kubernetes, and modern ML frameworks. Strong PyTorch and Python skills are essential.