Machine Learning Engineer in Bath

Contract length: 6 months (high chance of extension)

Outside IR35

Day rate: £450/day

Working model: Fully remote

Our client is a fast-scaling AI business building conversation intelligence technology, using AI to help large organisations get more value out of their customer and stakeholder conversations. They've already built serious credibility with some major household names in their sector. They're now pushing into genuinely greenfield territory, building out new platform capability and expanding into new markets overseas, while growing a team here in the UK to lead the charge.

This is a chance to come in as the ML subject matter expert and take control of most of the programming work within that build. You'll be bridging applied machine learning with proper platform engineering, taking existing models (NLP, generative AI, transcription) and turning them into standardised, production-ready components that actually hold up under real quality, cost and latency pressure.

What you'll be doing:

• Refactoring and upgrading existing ML models into modular, production-ready contracts
• Building resilient ML pipelines using DAG-based orchestration tools like Argo Workflows or Airflow, with proper retry logic and error handling
• Defining and automating evaluation pipelines using golden datasets so new models meet quality thresholds before release
• Designing provenance tracking for models, prompts and input data, so everything is auditable and reproducible
• Building out shadow testing and A/B testing infrastructure on live traffic, plus fallback and kill-switch mechanisms for safe rollout
• Engineering human-in-the-loop feedback pipelines to capture reviews and corrections and feed them back into training data
• Integrating third-party AI APIs and building out telemetry to track cost, token usage and latency

What we're looking for:

• Proven experience sitting across ML, MLOps and platform/backend engineering
• Strong understanding of evaluation metrics for generative AI, LLMs and/or speech models (Precision/Recall/F1, WER/CER, groundedness, hallucination rates)
• Solid hands-on experience with Docker, Kubernetes and modern orchestration frameworks
• Track record building observable ML systems with logging, monitoring and cost attribution
• Familiarity with data provenance, compliance standards and building fail-safe mechanisms for sensitive data and AI outputs

This is a proper hands-on build with real ownership, not a maintenance role, so if you want to be the one steering the technical direction rather than just executing someone else's, this is worth a conversation.

At Inara Talent, we believe everyone deserves a fair chance to shine. We connect great people from all backgrounds with opportunities where they can thrive, making sure hiring is fair, inclusive, and truly diverse. No matter your race, gender, age, ability, or background—we focus on what really matters: your talent.