Senior Architect (Agentic AI & Automations)

The Agentic AI Automations Architect is the senior technical authority for the design and delivery of conversational and agentic automations that resolve enterprise service requests across IT and Customer Service. Working on a modern Agentic AI Platform from our Partner, the Architect remains substantially hands-on, building and delivering to assigned customer accounts, while also owning the overall solution architecture, setting the patterns and standards the team builds to, and shaping solutions with customers from first conversation through to live production.

Key responsibilities:

• Architect and build conversational and process automations that collect from and act in external systems via platform connectors or direct API calls, covering knowledge retrieval, service request fulfilment, identity and access tasks, and guided troubleshooting, with safe handover to a live support experts.
• Build agentic automations driven by prompts and function libraries, defining the integrations, functions, and actions an automation can use so the generated prompt template selects the right function and populates its parameters from the request and connected data. The skill is in applying the LLM where it adds value while keeping flow logic explicit through designed nodes, branches, and decisions.
• Own reference architectures, patterns, and standards for the Agentic AI platform at expert level, translating each customer's context into the right design, judging which capabilities to apply for their environment, systems, constraints, and objectives, and curating proven, reusable patterns, flows, and components into a shared library the whole team designs and builds against.
• Own the solution architecture for assigned accounts, from solutioning through to live, supported production, engaging stakeholders directly to gather requirements, agree approach, and demonstrate value; owning design decisions and trade-offs and documenting them so the solution is defensible and maintainable; and overseeing build, deployment, and administration in the customer's environment, whether built personally or delivered with Engineers.
• Map and integrate the customer's system landscape, including systems of record (ITSM), identity directories, collaboration and productivity suites such as Microsoft 365, endpoint and digital-experience tooling, and CRM, ERP and HR systems, along with the data sources behind them, identifying the source of truth for each.
• Work fluently across hosting topologies, from public cloud (strong emphasis on Microsoft Azure) to private cloud, on-premises data centres, and hybrid combinations, and within the customer's network and connectivity boundaries.
• Design identity and access integration, connecting to the customer's identity provider for authentication, single sign-on, and federation with role-based, least-privilege access so automations act with auditable permissions and as the right user where needed.
• Act as design authority, deciding where each piece of logic and data should live, whether to use the platform, Azure, or a customer system for a given capability, and designing integrations resilient to the customer upgrading or replacing systems. Review the designs of Engineers, give clear technical direction, and keep solutions consistent with the team's reference architectures.
• Raise the capability of the team, mentoring Engineers, running design and code reviews, sharing reusable assets and good practice, and helping to evolve the way the CoE designs, builds, and delivers, all through influence rather than line management.
• Use Microsoft Azure to complement and extend the platform, filling gaps and fine-tuning where the platform alone is not sufficient. Apply Azure-native AI and agent services, in particular Microsoft Azure AI Foundry, its agent capabilities, model catalogue, and grounding and retrieval, both as an equivalent Agentic AI capability and a complement in Microsoft-heavy environments. Build supporting components (serverless functions, integration and workflow services, API gateways, event messaging) for orchestration, custom logic, and data preparation.
• Evaluate capabilities and inform the technical roadmap, assessing new Agentic AI and innovation features, Azure and Azure AI Foundry services, models, and emerging standards, running proofs of concept where useful; what to standardise on, and how the overall architecture should evolve.
• Embed enterprise security throughout, including zero-trust principles, secure handling of secrets and credentials, encryption in transit and at rest, PII masking, and content access controls per enterprise and regulatory requirements.
• Test, deploy, and continuously improve, with functional and reliability testing of real conversations and edge cases before production, promoting automations through development, test, and production environments under the customer's change process, and using analytics, unresolved-request review, and user feedback to increase automation and containment rates.

Required Skills and Experience:

• Substantial hands-on experience designing and building Agentic AI automations or integrations on an enterprise conversational AI or agentic AI platform, or strong transferable experience in workflow automation and systems integration, with a track record of owning solution architecture, not only building to someone else's design. This includes designing agentic automations where a model selects and calls functions while the flow logic stays explicit and controlled.
• Confident scripting for automation work, in particular JavaScript, with strong JSON handling (parsing, building, and transforming payloads) and data mapping between systems using approaches such as JSON-path mapping and JOLT-style transformations.
• Strong enterprise architecture experience, with proven ability to design solutions across real customer environments: networks, servers, core enterprise services, and the mix of cloud, on-premises, and hybrid hosting most organisations run, Microsoft infrastructure prominent among them. Comfortable producing reference architectures and articulating design decisions and trade-offs to technical and senior audiences.
• Experience integrating with enterprise business applications and systems of record (ITSM and ticketing such as ServiceNow, CRM, ERP, HR, etc), including reading and writing to them and understanding their data models, permissions, and constraints.
• Strong Microsoft Azure cloud expertise: identity with Microsoft Entra ID, networking (virtual networks, network security groups, private endpoints, firewalls, hybrid connectivity), and core compute, integration, data, and AI services. This includes using Azure to complement and extend the Agentic AI platform capabilities.
• Expertise working with Azure AI Foundry to build or ground agents and use its model catalogue, particularly in Microsoft-heavy environments.
• Strong integration skills across multiple mechanisms: REST APIs, webhooks, authentication and authorisation flows, including OAuth 2.0 in its common grant types, with token lifecycle and refresh handling, JSON, and event or message-driven patterns. Comfortable onboarding third-party APIs, with exposure to integration and middleware platforms.
• Experience deploying and operating integration components across hybrid and on-premises environments, including Windows-hosted services or containers with secure outbound connectivity to SaaS.
• Experience delivering conversational chat and voice channels into environments such as Microsoft Teams, embedded web chat.
• Practical LLM skills for enterprise use, including prompt engineering and prompt management, selecting and configuring hosted models, RAG and grounding in trusted data, intent classification and entity or value extraction, knowledge ontology and synonym tuning, and evaluating model quality and reliability.
• Experience shaping, scoping, and estimating engagements, running discovery with customers, advising on feasibility and approach, and supporting sales and pre-sales with solution outlines, architectures, and effort estimates that can be costed and committed to.
• Technical leadership and mentoring, acting as design authority, running design and code reviews, setting standards, and raising the capability of engineers through influence rather than formal line management.
• A high degree of autonomy and ownership, able to own architecture across multiple customer accounts in parallel, act as the senior technical point of escalation, and take responsibility for solution quality and outcomes.

Desirable:

• Prior experience in a solution architect, technical lead, or principal engineer role, ideally in a consultancy, managed-service, or multi-customer delivery setting.
• Experience with one or more leading agentic AI or conversational automation platforms like UiPath or AutomationAnywhere, with adaptability to transfer skills to a new toolset.
• Experience delivering in regulated or enterprise environments where governance, auditability, and access control matter, including Microsoft-heavy estates.
• Relevant Microsoft certifications, in particular Azure Solutions Architect, alongside Azure administration, Azure AI, identity, and security, or equivalent demonstrable experience.
• Emerging agent-interoperability standards such as the MCP, A2A, etc.