Manager - Microsoft Agent Builder - TC - FS

At EY, we’re all in to shape your future with confidence. 

We’ll help you succeed in a globally connected powerhouse of diverse teams and take your career wherever you want it to go. 

Join EY and help to build a better working world. 

  The opportunity

This is a hands-on delivery role for a Manager who designs, builds and deploys Microsoft AI and agentic solutions for financial services clients. The role owns delivery workstreams from design and use-case decomposition through configuration, build, integration, testing, pilot, deployment and adoption, working to the target architecture agreed with technical architecture.

The Manager will work with business leads, product owners, technology, data, security, risk, compliance and change teams. They will translate business processes and control requirements into working agent, automation and data designs, make practical platform and configuration decisions within the agreed architecture, and explain those decisions in plain language to business stakeholders.

The role requires current hands-on depth across Microsoft Copilot and agent technologies, the Power Platform, and the identity, data and governance services that make them safe in a regulated estate. It also requires the judgement to select the right boundary between Microsoft 365 Copilot, Copilot Studio, Cowork, Autopilots and Power Platform, and to recognise when a use case needs Azure AI Foundry, Azure OpenAI Service or custom Azure components and should be escalated for architectural design.

Strong relationships with Microsoft teams are important. The Manager will work day to day with the Microsoft Account team, Specialist Team Unit (STU) sellers and Global Black Belt (GBB) specialists to confirm platform fit, unblock technical issues, and support co-sell and co-build activity, solution showcases and follow-on pipeline growth.

Your key responsibilities

Solution design and use-case engineering

• Lead discovery across business journeys, processes, applications, data sources, integration touchpoints, entitlements and control requirements.
• Produce problem statements, design principles, future-state journeys, process maps, solution designs, agent specifications and role or operating-model impacts.
• Decompose prioritised use cases into agents, topics, tools, actions, workflows, connectors, knowledge sources, human approval points and operational controls.
• Recommend and justify platform choices across Microsoft 365 Copilot, Copilot Studio, Cowork, Autopilots and Power Platform, and escalate to solution architecture where a use case needs Azure AI Foundry, Azure OpenAI Service or custom Azure components.
• Prioritise use cases on value, feasibility, data readiness, risk and complexity, and build the business case across financial, customer, operational and risk outcomes.
• Define MVP scope, sequencing, release plans, non-functional expectations and the adoption approach that goes with them.

Build, integration and data

• Design, build and test custom Cowork, Autopilot, Copilot Studio agents, including topics, knowledge sources, actions, tools, triggers, channels, connectors and orchestration behaviour.
• Build Power Platform solutions across Power Automate, Power Apps, Power BI and Dataverse, applying environment strategy, solution packaging and application lifecycle management.
• Integrate agents and automations with systems of record using certified and custom connectors, Microsoft Graph, Dataverse and REST APIs, working within rate limits, shared service accounts and legacy interfaces, and escalating where no usable API exists.
• Configure governed grounding and retrieval over SharePoint, Dataverse, Microsoft Fabric and curated data products, managing retrieval scope, permissions trimming, data freshness and provenance.
• Implement approval, write-back and reconciliation controls for actions that update systems of record, including rejection and resumption paths.
• Design agent behaviour that fails safely, covering fallback responses, escalation to a human, and clear handling of out-of-scope or low-confidence requests.

Security, governance and responsible AI

• Build to secure-by-design and responsible-AI standards, covering identity propagation, least privilege, environment separation, secrets handling and correct connector and data scope.
• Working with our Cyber team, apply Microsoft Entra ID and Microsoft Purview controls to the solutions you build, including sensitivity labelling, data loss prevention, auditability and retention.
• Build human-in-the-loop checkpoints before client-facing activity, credit or risk decisions, and writes to systems of record, with evidence captured at each step.
• Configure logging and traceability for prompts, retrieved context, tool calls, model outputs, approvals, actions and exceptions.
• Prepare solution evidence for security, privacy, model-risk and compliance review, and take design and control approvals through to a decision.

Delivery, testing and production readiness

• Lead delivery workstreams that move designs into prototypes, proofs of concept, MVPs, pilots and scaled production services.
• Set up development environments, source control, solution packaging and the deployment pipeline that moves agentic solutions safely from sandbox to enterprise production.
• Test non-deterministic systems using golden datasets, prompt and tool-selection regression, adversarial and prompt-injection testing, sampled human review, and regular playbacks and demos.
• Run user acceptance, performance and operational-readiness testing, and agree monitoring, alerting, licensing, capacity, cost management and the support model at handover.
• Write and assure requirements, user stories, acceptance criteria, test cases, controls and adoption plans.
• Manage technical risks, dependencies, issues, scope changes and delivery quality, and keep the team focused on measurable business outcomes and tracked benefits.

Team leadership, adoption and Microsoft ecosystem

• Act as the day-to-day delivery lead for client stakeholders and present designs, progress and technical decisions to steering groups, design authorities and decision forums.
• Lead small teams of consultants, analysts, AI engineers and low-code developers, coach junior colleagues and set the standard for build and review quality.
• Drive adoption and enablement through training, live builds, demos, champion networks and practical governance guidance, so solutions are used rather than only delivered.
• Build strong day-to-day relationships with Microsoft Account, STU specialist seller and GBB technical teams, and bring them in early to confirm platform fit and unblock technical issues.
• Align solutions to Microsoft priority workloads and reference architectures, and contribute to joint account planning, client briefings, demonstrations, proposals and solution showcases.
• Manage engagement scope, budget, quality and timelines, and support proposals, pitches and follow-on work.

  Skills and attributes for success

• Strong hands-on delivery experience building AI, agentic, automation or low-code solutions, ideally in consulting and regulated financial services.
• Proven track record taking solutions end to end, from design through build and testing to a working product in production use.
• Ability to explain a solution design end to end, including agent behaviour, data flows, identity, controls, approval points and failure paths.
• Ability to distinguish between Microsoft 365 Copilot, Copilot Studio, Power Platform and pro-code solut