Job Opportunities API

About Us Skybound Wealth Management is a global financial advisory company with employees across the UK, USA, Switzerland, Cyprus, Spain and UAE. We provide tailored financial advice to international clients, supported by expert teams across wealth planning, compliance and operations. Role Overview We are seeking an experienced and technically capable AI Platform & LLMOps Engineer to take ownership of the reliability, performance and economics of our models, prompts, agents and AI services in production. The successful candidate will be responsible for the operational layer between model providers and the applications that consume them. This will include deployment, observability, evaluation pipelines, provider integrations, routing, quotas, release controls, cost optimisation and incident response. Working closely with our software engineering, data, cybersecurity and product teams, the AI Platform & LLMOps Engineer will ensure our AI services are observable, resilient, repeatable, secure and cost-controlled. This role requires someone who can operate and improve the underlying AI platform while also understanding how AI services are integrated into existing production applications and business workflows. Key Responsibilities AI Platform Operations Operate and maintain the Group’s model gateway and AI provider integrations. Manage authentication, model routing, fallback processes, rate limits, quotas and provider controls. Maintain controls relating to model and provider versions, deprecations and service changes. Ensure AI services are resilient, scalable and capable of supporting production applications and workflows. Work closely with software engineering teams to integrate AI services into existing and new applications. Monitoring Build and maintain dashboards and alerts covering token usage, cost, latency, error rates and service availability. Monitor quality and safety evaluations, tool failures and AI usage by person, team, product and use case. Implement distributed tracing, metrics and logging using OpenTelemetry or equivalent technologies. Define and maintain service-level objectives, operational thresholds and escalation processes. Identify emerging performance, reliability or cost issues before they affect users or business operations. Release Management Build prompt, agent and model release pipelines with appropriate version control and approval processes. Implement automated testing and evaluation as part of the AI release lifecycle. Manage canary deployments, controlled releases and rollback processes. Define and measure model accuracy, service quality and business value, rather than focusing solely on infrastructure performance and cost. Establish evaluation frameworks that measure whether AI services successfully complete the intended task, case, document or workflow. Monitor model performance and identify potential drift or deterioration in output quality. Cost Management & Optimisation Develop clear unit economics for AI services, progressing from cost per token to cost per successful task, case, document or workflow outcome. Monitor and report AI consumption and expenditure across teams, products and use cases. Optimise spend through model routing, caching, batching, prompt and context reduction, model selection and capacity planning. Balance cost, speed, accuracy and quality requirements when selecting and configuring AI services. Explain cost and quality trade-offs clearly to engineering, finance and executive stakeholders. Resilience & Incident Management Conduct load, resilience and red-team testing across AI services and supporting infrastructure. Develop and maintain operational playbooks for provider outages, model drift, data leakage, prompt injection and unexpected cost increases. Investigate and respond to incidents affecting AI