Who we are

About this role

What you’ll do

• Design and implement production machine-learning, generative-AI and justified agent-based services, including inference, feature processing, serving and integration.
• Productionise validated ML, generative-AI and agent-based solutions using approved Data Platform capabilities.
• Establish MLOps and LLMOps practices covering versioning, testing, automated evaluation, continuous delivery, deployment, rollback and environment promotion.
• Implement proportionate AI evaluation covering quality, safety, hallucination, bias, robustness and task performance.
• Own production monitoring and operational readiness across drift, latency, availability and cost.
• Establish appropriate guardrails, runbooks, incident-response processes and audit evidence for production AI services.
• Partner with the Lead Data Scientist and Data Engineers to translate validated scientific approaches into reliable production services.
• Evaluate build-versus-buy decisions and provider trade-offs, considering performance, security, portability, reliability, provider dependency and cost.
• Apply appropriate privacy, security, human oversight, explainability, safety and model-risk controls throughout the production lifecycle.
• Use AI-assisted engineering practices while maintaining human accountability, validation, traceability and production discipline.

Who you are

• Substantial practical experience in software, platform or ML Engineering, including production ML/AI delivery.
• Strong Python and software-engineering capability across APIs, automated testing, containers and continuous delivery.
• Experience with Vertex AI or equivalent managed ML platforms and model-serving services.
• Practical experience implementing MLOps, including model registries, versioning, deployment, rollback and monitoring.
• Relevant experience with generative-AI evaluation and LLMOps.
• Knowledge of responsible AI engineering, including privacy, security, human oversight, explainability, safety and model-risk controls.
• Strong understanding of APIs, containers, cloud services and production observability.
• An understanding of statistical and machine-learning concepts sufficient to identify and challenge invalid production assumptions.
• Sound production judgement when balancing reliability, latency, scalability, portability, provider dependency and cost.
• The ability to collaborate effectively with Data Scientists and Data Engineers to convert validated approaches into controlled production services.

Not a perfect fit?

What's in it for you

• The opportunity to engineer and productionise machine-learning, generative-AI and justified agent-based solutions across scalable digital services.
• Ownership across the ML/AI production lifecycle, including deployment, evaluation, observability, operational readiness and control.
• The opportunity to establish and mature MLOps and LLMOps delivery practices.
• Practical experience applying responsible AI engineering across quality, safety, hallucination, bias, robustness, privacy and security.
• The opportunity to influence technical decisions across model serving, tooling, provider selection and build-versus-buy choices.
• Collaboration with Data Science and Data Engineering specialists on the transition from validated approaches to controlled production services.
• Experience balancing real-world AI engineering considerations including reliability, latency, scalability, portability, provider dependency and cost.
• Flexible UK working across Newbury and Paddington.

What skills you will learn

• Production engineering for machine-learning and generative-AI systems.
• MLOps and LLMOps, including automated evaluation, model versioning, deployment, rollback and environment promotion.
• Production AI evaluation covering hallucination, bias, safety, robustness, quality and task performance.
• Cloud-based ML and AI engineering using Google Cloud Platform, Vertex AI or equivalent technologies.
• Model and AI-service observability across drift, latency, availability and cost.
• Engineering and operational controls for agent-based AI solutions.
• Responsible AI engineering covering security, privacy, explainability, human oversight and model-risk controls.
• Production incident response, runbook development, guardrails and audit evidence.
• Technical assessment of build-versus-buy, portability and AI-provider trade-offs.

VOIS Equal Opportunity Employer Commitment

Join Us

Alert

Follow us on social media and #StayConnected

• LinkedIn: https://www.linkedin.com/company/vois/
• Facebook: https://www.facebook.com/voisglobal
• Instagram: https://www.instagram.com/voisglobal/
• You can also chat with our employees to learn more about our projects: https://lnkd.in/dpkrcvR2