Overview
Location: Toronto, Ontario (Hybrid – 3 days/week onsite)
Contract Duration: 14 months, 40-hour work week
Skills: Python, MLOps, LLMOps, Azure Databricks, NIST AI RMF, MLFlow, RAG, AI governance, GitHub Actions, Terraform, Jenkins
Rate: $90-100/hr
Industry: Bank
DESCRIPTION
Applied Intelligence is scaling enterprise AI across Microsoft Foundry, Databricks, Copilot Studio, and a mixed model estate spanning Anthropic, Azure OpenAI, other frontier models, and open source models.
This hands-on engineering role partners with AI Governance and Risk to translate NIST AI RMF and enterprise requirements into practical platform controls, reusable patterns, automated guardrails, and governance evidence across the AI lifecycle.
-
- You’ll join the Applied Intelligence Platforms team and work closely with Platform Engineering, Data Engineering, Enterprise Architecture, Privacy, Enterprise Security, AI Governance and Risk, and the AI Centre of Excellence. You will report to the Senior Manager, Applied Intelligence Platforms.
-
- Apply safe deployment practices including automated builds, artifact versioning, deployment gates, and rollback strategies such as blue/green, canary, and shadow, avoiding manual production changes.
- Develop operational runbooks, support procedures, incident response guidance, and recovery processes.
Responsible AI, Privacy and Governance Enablement
-
- Partner with AI Governance, Privacy, Information Security, and the AI Centre of Excellence to translate enterprise requirements into practical implementation patterns.
- Implement technical guardrails supporting responsible AI outcomes, including content safety, grounding, human-in-the-loop checkpoints, and auditability.
- Apply secure AI implementation patterns aligned with OWASP guidance for LLM and agentic systems, mitigating risks such as prompt injection, insecure tool usage, model exfiltration, and data leakage.
- Implement authentication and authorization flows using OAuth2 and OIDC so AI agents and services inherit appropriate user context and role-based access.
- Implement privacy-supporting controls in AI solutions, such as data minimization, access scoping, retention alignment, and handling of sensitive data in prompts, logs, and retrieval sources.
AI delivery model
-
- Define and operationalize the AI delivery model, including reusable implementation patterns, reference solutions, and engineering standards for RAG, agentic, and machine learning solutions.
- Select and integrate models across the estate based on suitability, performance, and cost.
- Partner with delivery teams to accelerate build and support production readiness.
MLOps, LLMOps and evaluation
-
- Build and maintain pipelines covering development, evaluation, deployment, monitoring, and model lifecycle management.
- Establish evaluation frameworks with measurable quality criteria, behavioural regression testing, and continuous evaluation of probabilistic systems.
- Implement lineage, experiment tracking, and reproducibility using tools such as MLflow.
AI governance technical implementation
-
- Translate NIST AI RMF and enterprise AI governance requirements into technical controls across Govern, Map, Measure, and Manage.
- Implement reusable guardrails through platform configuration, CI/CD/IaC, evaluation gates, access controls, policy enforcement, and OWASP-aligned secure AI patterns.
- Establish traceable governance evidence for models, agents, deployments, evaluations, approvals, and human-in-the-loop controls.
- Partner with AI Governance and Risk to validate control effectiveness and escalate policy, privacy, security, and risk-acceptance decisions to accountable teams.
Operations, reliability and cost
-
- Own the reliability, observability, and production support of enterprise AI services.
- Own cost tracking, attribution, and optimization, treating cost as a measurable engineering metric.
- Apply safe deployment practices and maintain runbooks, incident response, and recovery processes.
- Enablement
- Onboard teams onto approved AI platforms and patterns, and provide technical mentorship, documentation, and reusable assets.
Decision making
-
- Independent analysis approach based on business context
What you’ll need:
Qualifications:
-
- Has shipped and supported production systems, and designs for failure.
- Has debugged performance or latency issues in distributed systems.
- Writes modular, testable code and favours automation over manual intervention.
- Defines measurable success criteria before building, and can test systems where outputs vary.
- Treats cost as a measurable engineering metric.
- Bachelor’s degree in a related technical field, or equivalent practical experience.
Knowledge:
-
- Generative AI, LLMs, RAG, agentic systems, and enterprise AI application patterns.
- Microsoft Foundry, Databricks/Unity Catalog/MLflow, Copilot Studio, AI Gateway/model access patterns, and mixed-model tradeoffs.
- MLOps/LLMOps, CI/CD, Infrastructure as Code, evaluation, observability, and cloud-native engineering.
- NIST AI Risk Management Framework (Govern, Map, Measure, Manage) and translation into engineering controls.
- Responsible AI, human-in-the-loop design, auditability, privacy, IP, security, bias, and model misuse risks.
- Cloud fundamentals, distributed systems and API design, observability, and data literacy.
- Cloud cost management, consumption tracking, and cost attribution.
- Proficiency in Python.
REQUIREMENTS
-
- Hands-on Python coding, AI, machine learning, platform, or cloud engineering.
- Hands-on experience building and operating production AI or machine learning solutions.
- Experience implementing RAG and agent-based solutions.
- Experience implementing MLOps and LLMOps capabilities end-to-end.
- Experience implementing automated AI governance controls or evidence, ideally mapped to NIST AI RMF or another recognized framework.
- Experience owning cost tracking and optimization of cloud or AI workloads.
- Experience in a regulated environment is an asset.
- Experience collaborating with Enterprise Architecture, Platform Engineering, Privacy, Security, and Governance teams in a large organization.
- Experience in financial services, pension, asset management, or another regulated environment is an asset.
NOTE: As part of our hiring process, we use AI-based systems to support the initial screening of applicants.
