AI Solution Architect

S.I. Systems Ltd.; Toronto; On-site; CAD 180,000 - 240,000; Full time

Job summary

S.I. Systems Ltd. in Toronto seeks a senior AI architect who owns end-to-end architecture for AI, generative AI and ML initiatives, from discovery to production operation.

You will work across business, technology, data, risk and vendor teams to craft governed, observable, resilient, cost-aware solutions aligned with enterprise roadmaps. This role collaborates with data scientists and engineers, validating material decisions with prototypes, prompt and retrieval tests, model evaluations, threat

Qualifications

• Typically 10+ years in technology with architecture responsibility and substantial production AI/ML delivery.
• End-to-end architecture ownership across models, data, retrieval, applications, integration, cloud, security and operations.
• Strong generative-AI knowledge: foundation models, RAG, embeddings, vector search, prompt orchestration, tool use and agents.
• Practical classical ML grounding: model lifecycle, feature and data pipelines, evaluation methods and common failure modes.
• Cloud AI services, model hosting, API and event integration, identity, networking, secrets and containers.
• Defining MLOps/LLMOps, evaluation, observability, release, rollback and model-monitoring requirements.
• Working knowledge of responsible AI, AI security threats, privacy engineering, data governance, model risk and human oversight.
• Ability to assess vendor and model evidence and compare hosted/open-weight/customized options on quality, risk, cost and control.
• Strong facilitation, stakeholder management and communication skills.
• Degree in computer science, engineering, data science or related field, or equivalent experience.

Responsibilities

• Own AI solution architecture from discovery through release and monitoring, providing assurance at experiment approval, production release, material model change and retirement.
• Translate business outcomes into AI tasks, decision boundaries, acceptance criteria and measurable value hypotheses.
• Select patterns (predictive ML, generative AI, RAG, agents, intelligent document processing or non-AI) and compare build, buy and consume options.
• Lead AI impact assessments covering data fitness, model risk, privacy, security, bias, explainability, IP, regulatory, cost and operating-model impacts.
• Define evaluation strategies and release thresholds for accuracy, relevance, groundedness, hallucination, toxicity, robustness, latency and unit economics.
• Manage AI-specific risks, dependencies, technical debt and vendor lock-in, keeping designs aligned with responsible-AI policy, approved platforms, data standards and roadmaps.
• Communicate benefits, limitations, residual risks and trade-offs clearly to technical, business and governance audiences.

Skills

End-to-end architecture

Generative AI knowledge

Data governance

Cloud & security

MLOps/LLMOps

Stakeholder management

Communication

Data science background

Education

Data science or related field

Bachelor’s degree in CS/Engineering or equivalent

Tools

Azure AI services

Azure Machine Learning

Azure OpenAI

Vector databases

Knowledge graphs

Job description

A senior individual contributor who owns end-to-end architecture for AI, generative-AI and machine-learning initiatives, from use-case discovery through production operation. Working across business, technology, data, risk and vendor teams, the architect selects the right AI pattern and turns it into a governed solution that is responsible, observable, resilient, cost-aware and aligned with enterprise roadmaps.

This is not a governance-only or conceptual role. The architect works closely with data scientists and engineers, validates material decisions with prototypes, prompt and retrieval tests, model evaluations, threat assessments and readiness checks, and knows when deterministic controls, human review or non-AI alternatives are needed.

Key accountabilities

• Own AI solution architecture from discovery through release and monitoring, providing assurance at experiment approval, production release, material model change and retirement.
• Translate business outcomes into AI tasks, decision boundaries, acceptance criteria and measurable value hypotheses.
• Select patterns (predictive ML, generative AI, RAG, agents, intelligent document processing or non-AI) and compare build, buy and consume options.
• Lead AI impact assessments covering data fitness, model risk, privacy, security, bias, explainability, IP, regulatory, cost and operating-model impacts.
• Define evaluation strategies and release thresholds for accuracy, relevance, groundedness, hallucination, toxicity, robustness, latency and unit economics.
• Manage AI-specific risks, dependencies, technical debt and vendor lock-in, keeping designs aligned with responsible-AI policy, approved platforms, data standards and roadmaps.
• Communicate benefits, limitations, residual risks and trade-offs clearly to technical, business and governance audiences.

Required qualifications

• Typically 10+ years in technology, with significant architecture responsibility and substantial production AI/ML delivery.
• End-to-end architecture ownership across models, data, retrieval, applications, integration, cloud, security and operations.
• Strong generative-AI knowledge: foundation models, RAG, embeddings, vector search, prompt orchestration, tool use and agents.
• Practical classical ML grounding: model lifecycle, feature and data pipelines, evaluation methods and common failure modes.
• Cloud AI services, model hosting, API and event integration, identity, networking, secrets and containers.
• Defining MLOps/LLMOps, evaluation, observability, release, rollback and model-monitoring requirements.
• Working knowledge of responsible AI, AI security threats, privacy engineering, data governance, model risk and human oversight.
• Ability to assess vendor and model evidence, and to compare hosted, open-weight, customized and vendor-embedded options on quality, risk, cost and control.
• Strong facilitation, stakeholder management and communication skills.
• Degree in computer science, engineering, data science or related field, or equivalent experience.

Preferred qualifications

• AI delivery in financial services, credit unions or another regulated industry.
• Azure AI services, Azure Machine Learning, Azure OpenAI or comparable, plus API management, containers and infrastructure as code.
• Model gateways, AI safety tooling, vector databases, knowledge graphs, feature stores, model registries and AI observability.
• Evaluating commercial AI products, embedded copilots, SaaS, managed model services and open-weight technologies.
• Enterprise architecture methods, recognized AI risk and governance frameworks, and relevant certifications.

A senior individual contributor who owns end-to-end architecture for AI, generative-AI and machine-learning initiatives, from use-case discovery through production operation. Working across business, technology, data, risk and vendor teams, the architect selects the right AI pattern and turns it into a governed solution that is responsible, observable, resilient, cost-aware and aligned with enterprise roadmaps.

This is not a governance-only or conceptual role. The architect works closely with data scientists and engineers, validates material decisions with prototypes, prompt and retrieval tests, model evaluations, threat assessments and readiness checks, and knows when deterministic controls, human review or non-AI alternatives are needed.

Key accountabilities

• Own AI solution architecture from discovery through release and monitoring, providing assurance at experiment approval, production release, material model change and retirement.
• Translate business outcomes into AI tasks, decision boundaries, acceptance criteria and measurable value hypotheses.
• Select patterns (predictive ML, generative AI, RAG, agents, intelligent document processing or non-AI) and compare build, buy and consume options.
• Lead AI impact assessments covering data fitness, model risk, privacy, security, bias, explainability, IP, regulatory, cost and operating-model impacts.
• Define evaluation strategies and release thresholds for accuracy, relevance, groundedness, hallucination, toxicity, robustness, latency and unit economics.
• Manage AI-specific risks, dependencies, technical debt and vendor lock-in, keeping designs aligned with responsible-AI policy, approved platforms, data standards and roadmaps.
• Communicate benefits, limitations, residual risks and trade-offs clearly to technical, business and governance audiences.

Required qualifications

• Typically 10+ years in technology, with significant architecture responsibility and substantial production AI/ML delivery.
• End-to-end architecture ownership across models, data, retrieval, applications, integration, cloud, security and operations.
• Strong generative-AI knowledge: foundation models, RAG, embeddings, vector search, prompt orchestration, tool use and agents.
• Practical classical ML grounding: model lifecycle, feature and data pipelines, evaluation methods and common failure modes.
• Cloud AI services, model hosting, API and event integration, identity, networking, secrets and containers.
• Defining MLOps/LLMOps, evaluation, observability, release, rollback and model-monitoring requirements.
• Working knowledge of responsible AI, AI security threats, privacy engineering, data governance, model risk and human oversight.
• Ability to assess vendor and model evidence, and to compare hosted, open-weight, customized and vendor-embedded options on quality, risk, cost and control.
• Strong facilitation, stakeholder management and communication skills.
• Degree in computer science, engineering, data science or related field, or equivalent experience.

Preferred qualifications

• AI delivery in financial services, credit unions or another regulated industry.
• Azure AI services, Azure Machine Learning, Azure OpenAI or comparable, plus API management, containers and infrastructure as code.
• Model gateways, AI safety tooling, vector databases, knowledge graphs, feature stores, model registries and AI observability.
• Evaluating commercial AI products, embedded copilots, SaaS, managed model services and open-weight technologies.
• Enterprise architecture methods, recognized AI risk and governance frameworks, and relevant certifications.

Disclaimer:
AI may be used in evaluating candidates.
This posting is for an existing vacancy.