1 to 25 of 250 Permanent Human-in-the-Loop Jobs in the UK

governance, ethics, compliance, and risk control activities. Ensure AI solutions comply with enterprise architecture principles, security policies, data governance standards, human-in-the-loop controls, and regulatory requirements. Monitor and mitigate AI risks, raising concerns early and recommending remedial actions. REQUIREMENTS To perform this job successfully ...

multi-agent systems that reason, plan, and act using tools. Command of agentic design patterns (e.g., ReAct, planning, reflection, tool use, human-in-the-loop) and agent frameworks such as LangGraph, CrewAI, MAF, the OpenAI Agents SDK, and Google’s Agent Development Kit (ADK). ...

tuning and model customisation where appropriate Agentic AI Multi-agent systems Tool calling Function orchestration Workflow automation Planning and reasoning patterns Human-in-the-loop design Knowledge & Retrieval RAG architectures Vector databases Semantic search Enterprise search Document ingestion Knowledge graphs (advantageous) Integration APIs Event-driven architectures ...

reliability, scalability, observability and ongoing support.Monitor agent performance, drift, hallucination risk, unintended actions, incidents and service degradation.Implement appropriate controls, guardrails, human-in-the-loop mechanisms and auditability across autonomous solutions.Contribute to reusable AI and automation standards, architecture patterns and engineering best practices.Communicate solution designs, outcomes, risks ...

governance, model risk, information security, data privacy, records management, regulatory, compliance and operational resilience requirements.Design controls into AI-enabled processes, including human-in-the-loop review, explainability, validation, testing, monitoring, exception handling, evidence capture and audit trails.Collaborate with Global Markets AI and the ...

Agentic RAG, Graph RAG, knowledge-graph, and hybrid approaches. Design multi-agent systems with modern orchestration and reasoning architectures, including human-in-the-loop and autonomous frameworks. Define memory, context, planning, tool-use, and reasoning strategies for agentic systems. Stay technical. Contribute to architecture, critical-path ...

design. Agentic AI & Multi‐Agent Architecture – design of autonomous and multi‐agent systems, agentic design patterns (ReAct, planning, reflection, tool use, human‐in‐the‐loop), and frameworks (LangGraph, CrewAI, MAF, OpenAI Agents SDK, Google ADK). Proven ability to architect reliable agentic workflows with memory, state ...

day. Responsible AI in Practice Implement the guardrails the Architect designs - bias mitigation checks, evaluation harnesses, human-in-the-loop review points - so that Responsible AI principles (ATRS alignment, NCSC "Secure by Design, " meaningful human control) are enforced in the ...

versioning, rollout/rollback), guardrails, telemetry, drift monitoring, prompt safety.DevSec & Responsible AI: Privacy‐by‐design, PII handling, secrets management, policy enforcement, human‐in‐the‐loop approvals, auditability.Languages: Advanced Python plus one of Java/TypeScript/Go; containerization (Docker/K8s).Leadership & Communication:Executive presence: influence ...

selection, and evaluation tradeoffs.Familiarity with applied AI patterns such as RAG, embeddings, vector search, document extraction, tool calling, agentic workflows, and human-in-the-loop review.Strong programming and data skills, including Python, SQL, and tools/libraries used in agentic development like langgraph, langchain, Pydantic ...

platforms, or CI/CD for ML and LLM systems. Experience implementing responsible-AI or AI-governance controls such as guardrails, human-in-the-loop oversight, and audit trails in production. Experience mentoring engineers or setting technical standards as a senior individual contributor. What ...

technical audiences.Excellent understanding of model validation, feature engineering, explainability, data quality and governance, along with an appreciation of responsible AI practices, human-in-the-loop decision making and appropriate AI controls.A curious mindset with a willingness to adopt approved AI-enabled workflows and automation to improve ...

model risk, information security, data privacy, records management, regulatory, compliance and operational resilience requirements. Design controls into AI-enabled processes, including human-in-the-loop review, explainability, validation, testing, monitoring, exception handling, evidence capture and audit trails. Collaborate with Global Markets AI and the ...

outputs are accurate, grounded, and commercially reliable Embedding AI governance, observability, and responsible AI principles into architecture design — including audit logging, human-in-the-loop escalation points, and performance monitoring — as first-class concerns rather than afterthoughts Providing technical leadership to delivery teams — guiding engineers, reviewing ...

point of view grounded in delivery experienceEmbedding responsible AI and governance principles into agentic delivery programmes — including auditability, explainability, and human-in-the-loop mechanismsContributing to Accenture Song's commerce delivery capability — including shaping accelerators, reusable assets, delivery standards, and points of view that advance ...

platforms, or CI/CD for ML and LLM systems. Experience implementing responsible-AI or AI-governance controls such as guardrails, human-in-the-loop oversight, and audit trails in production. Experience mentoring engineers or setting technical standards as a senior individual contributor. What ...

what the team has not thought about yetGenuine familiarity with agentic AI patterns — multi-agent orchestration, tool use, memory management, human-in-the-loop design — and where they succeed and fail in real delivery contextsRequired — Technical LeadershipExperience leading engineering and cross-functional teams ...

including quality measurement, hallucination reduction, regression testing, model performance tracking, and operational risk controlsDesign approaches for continual learning and improvement, including human-in-the-loop feedback, telemetry-driven enhancement, model, prompt, and version management, and safe release practicesCollaborate with Markets Operations stakeholders to understand process pain ...

augmentation incident response strategy for security operations: agentic AI workflows, LLM-assisted triage, prompt libraries, notebook-based investigation, and human-in-the-loop autonomous response patterns.• Establish governance, guardrails, and auditability for AI-assisted and AI-autonomous response actions in support of regulatory frameworks ...

knowledge of prompt engineering and evaluation techniques for agent reliability, safety, and cost control. Familiarity with orchestrating multi-agent workflows and human-in-the-loop approval steps for sensitive actions. Awareness of security, data-privacy, and governance considerations specific to AI agents (prompt injection risks, access ...

solution design, including model selection, prompt and context design, grounding, tool use, structured outputs and scalability trade-offs.Agentic AI, orchestration and human in the loop patterns.Classical and applied machine learning architecture, including supervised and unsupervised learning, predictive models, recommendation systems, similarity models, optimization, model validation, feature ...

agentic platform: orchestration patterns (supervisor/sub-agent), agent decomposition, tool-calling design, memory and state management, error handling and human-in-the-loop integration. Lead implementation across the agent framework stack — LangGraph, LangChain or equivalent — with a strong opinion on framework choice and lifecycle. ...

Strong understanding of how to consume AI within workflows, including LLM calls for classification, extraction and summarisation, intelligent document processing and human-in-the-loop design. Practical judgement on when to use AI in automation, and when deterministic, rules-based automation is safer, simpler ...

best practicesCreate durable, reusable software frameworks and patterns leveraged across teamsDesign and govern agentic AI systems, including multi-agent workflows and human-in-the-loop controlsEstablish engineering standards for LLM-based applications, ensuring safety and reproducibilityDrive adoption of advanced technical methods and industry-aligned product development ...

Create durable, reusable software frameworks and patterns leveraged across teams Design and govern agentic AI systems, including multi-agent workflows and human-in-the-loop controls Establish engineering standards for LLM-based applications, ensuring safety and reproducibility Drive adoption of advanced technical methods and industry-aligned ...