21 of 21 Remote/Hybrid Model Context Protocol Jobs in South Wales

role is engineering AI agents and agentic workflows. You will design how agents use trusted knowledge, approved tools, Model Context Protocol (MCP) services, approvals, human hand-offs and safe failure paths. You will ensure that evaluation, observability, identity, permissions and data boundaries are built into delivery rather … agents and agentic workflows using trusted knowledge, tools, actions, approvals, human hand-offs, evaluation and safe failure paths. • Design and implement integrations with approved MCP servers and lead development of MCP tools or services using secure, reusable patterns. • Integrate Power Platform and AI solutions with Microsoft 365, SharePoint, Microsoft Graph ...

admin position. A. Agentforce Architecture & System Design Solution Architecture: Translate complex enterprise business processes into end-to-end agentic workflows (Intent \rightarrow Agent \rightarrow Context \rightarrow Reasoning \rightarrow Actions \rightarrow Systems \rightarrow Outcome). Core Agent Engineering: Design, configure, and build enterprise agents using Agent Builder, Agentforce Studio, Agent … search, and Knowledge libraries. System Interoperability: Architect REST API callouts, event-driven integrations, and MuleSoft platform pipelines, incorporating Model Context Protocol (MCP) standards. D. Testing, Security & Strategic Advisory Testing & Quality Assurance: Drive automated testing and multi-turn conversation tracing using the Agentforce Testing Center. Security & Guardrails: Design ...

team and set the bar in design reviews and code reviews. Engage with the open source and agentic development community around sandboxing, MCP, and related ecosystems. Participate in on-call rotations and respond to critical stability issues affecting users. Document your work clearly and share knowledge with the rest ...

layer the protocols and flows that determine how every Grafana user proves who they are. We actively implement cutting-edge authentication protocols such as MCP OAuth (OAuth 2.1), OIDC, SAML, LDAP, and SCIM, rather than relying on off-the-shelf components. We are completely redesigning Grafana s authentication ...

layer — the protocols and flows that determine how every Grafana user proves who they are. We actively implement cutting-edge authentication protocols such as MCP OAuth (OAuth 2.1), OIDC, SAML, LDAP, and SCIM, rather than relying on off-the-shelf components. We are completely redesigning Grafana’s authentication ...

performance of sandbox creation, session management, and teardown for AI agent workloads. Engage with the open source and agentic development community around sandboxing, MCP, and related ecosystems. 8+ years of software engineering experience, with demonstrated technical leadership on production systems at scale. Bachelor s degree in Computer Science, Engineering ...

performance of sandbox creation, session management, and teardown for AI agent workloads. Engage with the open source and agentic development community around sandboxing, MCP, and related ecosystems. 8+ years of software engineering experience, with demonstrated technical leadership on production systems at scale. Bachelor s degree in Computer Science, Engineering ...

agent can use?" before asking "does this data exist?". As that foundation takes shape, your scope expands naturally into Grip's MCP design, agentic workflows and AI system architecture. In your own workflow, agentic tooling is already the default. You use Claude Code, GitHub automations and agents to manage ...

expand into agentic solution delivery and broader business transformation. Our conviction is that engineers are uniquely positioned to lead this transformation. The strongest mental model for designing, deploying, and governing AI agents in production lives in software engineering. The teams that change how their organization builds software … assumptions, risks, milestones, and dependencies Embed with customer teams to co-build agentic applications on the Claude developer platform Build technical artifacts evaluation frameworks, MCP servers, sub-agents, and deployment configurations tailored to the customer's stack What Success Looks Like Customers move from proof-of-value to production Engineering ...

expand into agentic solution delivery and broader business transformation. Our conviction is that engineers are uniquely positioned to lead this transformation. The strongest mental model for designing, deploying, and governing AI agents in production lives in software engineering. The teams that change how their organization builds software … assumptions, risks, milestones, and dependencies Embed with customer teams to co-build agentic applications on the Claude developer platform Build technical artifacts evaluation frameworks, MCP servers, sub-agents, and deployment configurations tailored to the customer's stack What Success Looks Like Customers move from proof-of-value to production Engineering ...

build pipelines Software supply chain integrity Container security and governance AI/ML workload governance (where applicable) AI agent architecture patterns: isolated execution environments, MCP policy enforcement, model artifact provenance Partner with Sales Engineers and Post-Sales teams to ensure technical continuity without owning deal cycles. What … packaging, dependency management, reproducibility, and governance considerations within containerized environments. Practical Experience with AI coding agent runtimes and sandboxed execution environments (e.g., agent isolation, MCP tool governance) Familiarity with Docker SBX and MCP Gateway as enterprise controls for agentic AI workloads Familiarity with secure software supply chain practices, including SBOM ...

build pipelines - Software supply chain integrity - Container security and governance - AI/ML workload governance (where applicable) - AI agent architecture patterns: isolated execution environments, MCP policy enforcement, model artifact provenance - Partner with Sales Engineers and Post-Sales teams to ensure technical continuity without owning deal cycles. What … packaging, dependency management, reproducibility, and governance considerations within containerized environments. - Practical Experience with AI coding agent runtimes and sandboxed execution environments (e.g., agent isolation, MCP tool governance) - Familiarity with Docker SBX and MCP Gateway as enterprise controls for agentic AI workloads - Familiarity with secure software supply chain practices, including SBOM ...

Code to entire teams shipping together with AI-native transformation. Our conviction is that engineers are uniquely positioned to lead this transformation. strongest mental model for designing, deploying, and governing AI agents in production lives in software engineering. software are the same teams equipped to shepherd the rest … changes existing engineering practices: issue management, branching, testing, code review, and security Scope engagements: assumptions, risks, milestones, and dependencies Build technical artifacts evaluation frameworks, MCP servers, sub-agents, and deployment configurations tailored to the customer's stack Engineering teams measurably change how they ship software Accounts expand on the strength ...

moving from a traditional SaaS posture to a platform built around agents - parent agents fronting our compensation data, specialist agents underneath, durable workflows, an MCP-based tool layer, and an eval harness to maintain consistency. You ll own architecture and technical design across the stack: the customer-facing … have Production experience with agentic systems - tool calling, multi-step orchestration, LLM evals. Workflow engines (Temporal, Cadence, Step Functions) - even if not Temporal specifically MCP, Anthropic SDK, or other LLM platform work Building developer platforms, internal SDKs, or framework-shaped surfaces Eval-driven development, prompt engineering at scale, LLM-judge ...

moving from a traditional SaaS posture to a platform built around agents - parent agents fronting our compensation data, specialist agents underneath, durable workflows, an MCP-based tool layer, and an eval harness to maintain consistency. You ll own architecture and technical design across the stack: the customer-facing … have Production experience with agentic systems - tool calling, multi-step orchestration, LLM evals. Workflow engines (Temporal, Cadence, Step Functions) - even if not Temporal specifically MCP, Anthropic SDK, or other LLM platform work Building developer platforms, internal SDKs, or framework-shaped surfaces Eval-driven development, prompt engineering at scale, LLM-judge ...

agent can use?" before asking "does this data exist?". As that foundation takes shape, your scope expands naturally into Grip's MCP design, agentic workflows and AI system architecture. In your own workflow, agentic tooling is already the default. You use Claude Code, GitHub automations and agents to manage … Define success metrics and build feedback loops that keep product decisions grounded in data. Expanding into AI & agentic - as your scope grows Contribute to MCP tool design, shaping how Grip's data is exposed and consumed within agentic systems. Work with the Data and AI squad to identify where agentic ...

moving from a traditional SaaS posture to a platform built around agents - parent agents fronting our compensation data, specialist agents underneath, durable workflows, an MCP-based tool layer, and an eval harness to maintain consistency. You ll own architecture and technical design across the stack: the customer-facing … React or another modern frontend framework Frontend React Typescript Typescript PostgreSQL GCP (Cloud Run, Cloud SQL, Cloud Storage + more) AI/agentic platform MCP servers as the tool layer AI gateway (Bifrost) for logging, routing, governance From your perspective, there are three critical things to look for when evaluating ...

moving from a traditional SaaS posture to a platform built around agents - parent agents fronting our compensation data, specialist agents underneath, durable workflows, an MCP-based tool layer, and an eval harness to maintain consistency. You ll own architecture and technical design across the stack: the customer-facing … React or another modern frontend framework Frontend React Typescript Typescript PostgreSQL GCP (Cloud Run, Cloud SQL, Cloud Storage + more) AI/agentic platform MCP servers as the tool layer AI gateway (Bifrost) for logging, routing, governance From your perspective, there are three critical things to look for when evaluating ...

agent can use?" before asking "does this data exist?". As that foundation takes shape, your scope expands naturally into Grip's MCP design, agentic workflows and AI system architecture. This position is a remote role based in the UK. Building the data layer for agentic systems - immediate focus Diagnose … Define success metrics and build feedback loops that keep product decisions grounded in data. Expanding into AI & agentic - as your scope grows Contribute to MCP tool design, shaping how Grip's data is exposed and consumed within agentic systems. Work with the Data and AI squad to identify where agentic ...

agent can use?" before asking "does this data exist?". As that foundation takes shape, your scope expands naturally into Grip's MCP design, agentic workflows and AI system architecture. This position is a remote role based in the UK. Building the data layer for agentic systems - immediate focus Diagnose … Define success metrics and build feedback loops that keep product decisions grounded in data. Expanding into AI & agentic - as your scope grows Contribute to MCP tool design, shaping how Grip's data is exposed and consumed within agentic systems. Work with the Data and AI squad to identify where agentic ...

beyond initial onboarding. Structured PoCs experience producing technical blogs or deep technical content. Preferred Hands-on experience with AI/ML development workflows, including model … packaging, dependency management, reproducibility, and governance considerations within containerized environments. Practical Experience with AI coding agent runtimes and sandboxed execution environments (e.g., agent isolation, MCP tool governance) Familiarity with Docker SBX and MCP Gateway as enterprise controls for agentic AI workloads Familiarity with secure software supply chain practices, including SBOM ...