Lead AI Solutions Architect and Engineer, Microsoft Copilot
Position Overview:
-
Manager is trying to create a high-level CoPilot feature to pull analysis from various dashboards.
- This group is on the business side of Schwab; think Asset Management. Not engaging directly with the customers.
- They are using CoPilot to pull data from Salesforce, inboxes, dashboards to feed the manager all the information they need regarding pipelines, gaps, and portfolios.
We are seeking a hands-on Lead AI Solutions Architect and Engineer to lead the discovery, architecture, and development of enterprise AI agents and intelligent workflow solutions using Microsoft Copilot technologies.
This individual will work directly with business and technology stakeholders to understand use cases, workflows, data sources, access constraints, and expected outcomes. They will translate those findings into a practical architecture and phased implementation plan, while also developing agents, integrations, prototypes, and reusable technical components alongside the delivery engineer. This is not a strategy-only or documentation-focused architect role. The successful consultant must be equally comfortable facilitating discovery, making architectural decisions, configuring and extending Copilot Studio, integrating enterprise systems, troubleshooting technical issues, and contributing directly to production-ready solutions.
The initial use cases involve consolidating branch intelligence currently distributed across reporting sources and creating an agent that converts dictated sales notes into structured, role-appropriate output. The broader objective is to establish reusable agent, data, integration, and workflow patterns that can support additional use cases over time.
What You’ll Do:
- Lead business and technical discovery sessions to understand use cases, personas, business processes, required outputs, data sources, access requirements, dependencies, and constraints.
- Partner with business stakeholders to translate loosely defined ideas into clear objectives, functional requirements, technical requirements, user stories, and acceptance criteria.
- Evaluate proposed use cases based on business value, technical feasibility, data readiness, integration complexity, security, and implementation risk.
- Identify relationships among use cases and define reusable agent, workflow, knowledge, integration, and data-access patterns.
- Design the overall architecture for Microsoft Copilot Studio agents, intelligent workflows, enterprise integrations, and supporting services.
- Define MVP boundaries, technical assumptions, implementation sequencing, dependencies, risks, and a phased roadmap.
- Build and configure Copilot Studio agents, topics, instructions, actions, knowledge sources, variables, and orchestration logic.
- Develop prototypes and technical spikes to validate architecture decisions and reduce implementation risk.
- Create secure agent actions and workflow automations using Power Automate, APIs, connectors, and supporting Microsoft technologies.
- Design and develop integrations with enterprise applications, reporting environments, databases, Microsoft 365 services, and approved data sources.
- Apply prompt and context engineering, grounding, retrieval, validation, and structured-output techniques to improve agent accuracy and consistency.
- Design guided questioning and validation logic that identifies incomplete input and prompts users for required information.
- Establish reusable components, connectors, templates, technical patterns, and development standards that can support multiple agents and use cases.
- Evaluate identity, permissions, role-based access, data entitlements, privacy, security, compliance, and responsible-AI requirements.
- Define exception handling, fallback processes, monitoring, observability, and operational-support requirements.
- Participate directly in development, testing, debugging, deployment, and production stabilization.
- Conduct technical design reviews and code or configuration reviews with the engineering team.
- Guide the solution from discovery through implementation rather than transitioning out after the architecture phase.
- Communicate architecture decisions, risks, tradeoffs, progress, and recommendations clearly to both technical and nontechnical stakeholders.
Required Qualifications:
- Significant experience designing, developing, integrating, and supporting enterprise applications or automation solutions.
- Hands-on experience building solutions with Microsoft Copilot Studio, Microsoft 365 Copilot extensibility, generative AI, or comparable enterprise agent technologies.
- Demonstrated ability to lead discovery while also contributing directly to development and implementation.
- Strong understanding of enterprise solution architecture, software design patterns, integration architecture, security, scalability, reliability, and maintainability.
- Experience designing and implementing APIs, custom connectors, workflow automations, and integrations with enterprise systems.
- Experience with Microsoft Power Platform, particularly Power Automate.
- Understanding of AI-agent concepts including instructions, knowledge sources, tools, actions, orchestration, grounding, and multistep workflows.
- Experience with prompt and context engineering, retrieval techniques, structured outputs, evaluation, testing, and guardrails.
- Experience assessing identity, authentication, authorization, role-based access, data permissions, and security constraints.
- Strong understanding of SDLC, Agile delivery, source control, environment management, testing, deployment, and production support.
- Ability to translate ambiguous business requirements into an actionable architecture, backlog, and implementation approach.
- Strong client-facing communication, facilitation, documentation, and stakeholder-management skills.
- Experience delivering solutions within a large, complex, or highly regulated enterprise environment.
- Ability to work effectively in a small delivery team where responsibilities may span discovery, architecture, development, testing, and troubleshooting.