AI Solution Architect

Purpose of the Job 

The AI Solution Architect is a senior individual-contributor member of the Solutions Architecture team, responsible for designing, documenting, and delivering solution-level architecture for initiatives that require AI, agentic, and machine learning capabilities. The role's focus is translating business and technical requirements into concrete, implementable solution designs — application architecture, integration patterns, data flows, and technology selection — for specific programs and projects, rather than setting enterprise-wide architecture strategy or standards. 

The incumbent carries a solution architecture caseload alongside their peers, designing AI/agentic solutions that integrate with existing systems across initiatives such as PC Financial (PCF) integration, the Enterprise Customer Information File (eCIF), and AML/Risk/Finance/Treasury delivery. The role partners with Enterprise Architecture, Enterprise Data Governance, the AI Centre of Enablement (AI CoE), and Cyber Security to ensure solutions comply with enterprise standards and are secure, scalable, and fit for purpose — while remaining focused on solving the specific technical problem in front of them rather than owning those enterprise standards.  

The role designs AI/agentic solutions primarily on Microsoft Azure, with working exposure to Google Cloud Platform (GCP) where initiatives require it (e.g., Vertex AI / Gemini Enterprise Agent Platform workloads), and applies open-source orchestration frameworks such as LangGraph/LangChain where they are the appropriate tool for the solution.

Solution Architecture & Delivery (core focus) 

Design and document solution architecture artifacts — component diagrams, integration diagrams, sequence/data-flow diagrams, and technical design specifications — for assigned initiatives, ensuring designs are implementable, secure, and aligned to approved enterprise standards. 

Translate business requirements into technical solution designs through requirements-gathering sessions and joint working sessions with business analysts, developers, and product owners. 

Select appropriate technology components, integration patterns (API, event-driven, batch), and cloud services to meet solution requirements, evaluating trade-offs between cost, performance, and maintainability. 

Present solution designs and technology intake requests to governance forums (e.g., Technical Review Body/Guild, Architecture Design Authority) for review and approval, incorporating feedback into final designs. 

Support delivery across concurrent priority programs — including PCF integration, eCIF, and AML/Risk/Finance/Treasury initiatives — acting as the named solution architect accountable for the technical design of assigned work packages. 

AI & Agentic Solution Design (specialized focus) 

Design and document end-to-end solution architecture for AI/agentic use cases (e.g., coded agents built with LangGraph/LangChain, Azure AI Foundry-hosted agents, GCP Vertex AI Agent Builder/Agent Engine-hosted agents, RAG patterns, APIM or Apigee AI gateways), including component design, network isolation, private endpoints, managed identity authentication, token governance, and observability. 

Design and implement stateful, multi-step agent workflows using the LangGraph/LangChain ecosystem (LangGraph for graph-based orchestration and durable execution, LangChain for tool/retriever/memory components, LangSmith for tracing and evaluation), applying human-in-the-loop and checkpointing patterns for production reliability. 

Design data access patterns for AI solutions (sources, ingestion, storage) that comply with enterprise data architecture standards, working with the Enterprise Data & Analytics team to validate approach across both Azure (Fabric/Synapse) and GCP (BigQuery) data platforms where applicable. 

Partner with the AI CoE, Cyber Security, and IAM teams on solution-level RBAC/persona models, guardrails, and promotion path (sandbox → pilot → production) for each assigned AI solution. 

Evaluate and prototype (proof-of-concept) emerging AI vendors, models, and tools for specific use cases, documenting solution-level recommendations and risks. 

Identify and implement practical AI/agentic automation opportunities within owned solution designs and delivery workflows (e.g., decision-record drafting, diagram generation) to improve delivery efficiency. 

Collaboration & Technical Leadership 

Provide technical guidance, consultation, and mentorship to developers, business system analysts, and fellow architects on solution-level design questions. 

Document solution architecture decisions, processes, and standards for owned initiatives; maintain accurate inventory and vendor records for solutions delivered. 

Collaborate with external consulting partners and technology vendors (e.g., Microsoft) on joint solution delivery engagements and semantic modelling workshops as they relate to assigned solutions. 

Contribute to departmental activities, on-call/production support for owned solutions, and other related tasks as required by the Director. 

Bachelor's degree in Computer Science, Computer Engineering, or a related field; a post-graduate degree is an asset. 

8+ years of progressive IT experience, including at least 5 years designing and delivering solution, application, or integration architecture for specific projects/initiatives, operating as a senior individual contributor without direct reports. 

Demonstrated, hands-on experience designing solution architecture — application, integration, and data solutions — for individual business initiatives, including experience moving solutions from design through implementation. 

Hands-on experience architecting AI and agentic solutions on Microsoft Azure (Azure AI Foundry, Azure OpenAI, APIM AI gateways, vector stores/AI Search) and/or Google Cloud Platform (Vertex AI, Gemini Enterprise Agent Platform / Agent Builder, Agent Engine, BigQuery), including related landing-zone patterns (networking, private endpoints/VPC-SC, managed identity/IAM). 

Practical experience with modern agent orchestration frameworks, particularly the LangGraph/LangChain ecosystem (LangGraph for stateful multi-agent orchestration, LangChain for composable components, LangSmith for observability), and familiarity with interoperability standards such as Model Context Protocol (MCP) and Agent2Agent (A2A) protocol for connecting agents across frameworks/clouds. 

Working knowledge of Responsible AI, model governance, and AI risk/compliance considerations (e.g., OSFI B-13/E-23 overlay) sufficient to design compliant solutions and support security/architecture reviews. 

Strong knowledge of microservices, RESTful, event-driven, and cloud-native architecture patterns, and experience with relational and NoSQL database technologies for transactional and analytical workloads. 

Working knowledge of data platform and analytics tooling (e.g., Microsoft Fabric, Synapse, Power BI) sufficient to design solutions that integrate cleanly with the Bank's data platform. 

Preferred certifications: Microsoft Certified: Azure Solutions Architect Expert, Azure AI Engineer Associate, or equivalent, Google Cloud Professional Machine Learning Engineer or equivalent. 

Experience in the banking or financial services industry, including familiarity with Canadian regulatory expectations, is strongly preferred. 

Excellent problem-solving skills, with the ability to design innovative, implementable solutions to ambiguous, project-specific technical challenges. 

Communication skills required are moderately complex to complex: providing technical information and solution recommendations, conducting/leading design working sessions, making presentations at governance forums, and producing clear technical documentation for both technical and business audiences.