The Opportunity:The AI Engineer reports to the AI Engineering Lead and works within First National's Residential Underwriting Product Squad, accountable for designing, building, testing, and improving AI enabled mortgage workflows that enhance productivity, decision quality, policy adherence, and user experience.
This is a hands on engineering role focused on agentic workflows, document and policy intelligence, retrieval augmented generation, deterministic rules and calculations, tool integration, human in the loop review, and measurable model quality.
How you will contribute:- Design and build AI enabled workflows for residential mortgage use cases, including document extraction, policy interpretation, retrieval augmented generation, summarization, tool calling, deterministic decision logic, calculations, exception handling, and human in the loop review.
- Translate First National business rules, underwriting policies, process steps, and edge cases into modular prompts, agents, retrieval pipelines, tools, services, and testable acceptance logic, in partnership with the Residential Underwriting Business SME and the Merlin Product Owner.
- Build on shared AI Factory services and approved patterns across Databricks and Azure, including governed data, Vector Search, model and prompt registries, evaluation harnesses, APIs, identity, logging, monitoring, and controlled access to model providers, and raise platform gaps to the AI Engineering Lead and the Technical Product Manager rather than building parallel platform components.
- Develop production quality code using Python, SQL, FastAPI, Git based workflows, automated tests, CI/CD, and modern engineering practices, and integrate AI services with Merlin, core mortgage systems, document repositories, data sources, and user facing applications through the interfaces and data contracts owned by the AI Data and Integration Engineer.
- Create representative use case test cases and evaluation datasets to the methodology, acceptance thresholds, and quality standards set by the AI Quality and Evaluation Lead, and work with IT QA to analyze failure patterns and improve grounding, policy adherence, confidence, precision, recall, and exception routing.
- Implement safeguards for privacy, security, auditability, model and prompt versioning, observability, feedback capture, rollback, and cost control, following First National engineering and production readiness standards.
- Support UAT, parallel runs, defect triage, production releases, incident analysis, and ongoing monitoring, and rapidly address edge cases identified through user feedback, evaluations, and operational telemetry.
- Refactor partner delivered or prototype components into modular, documented, maintainable services, and contribute reusable prompts, tools, retrieval patterns, and lessons learned back to the AI Factory component catalog.
The experience you need:- Bachelor's degree in computer science, engineering, data science, mathematics or a related discipline.
- 5 plus years of progressive software engineering, machine learning engineering, data engineering, applied AI, or related technology experience.
- Hands on experience building production AI or generative AI solutions, including RAG, agentic workflows, document intelligence, prompt and tool design, model serving, evaluation, observability, and human in the loop controls.
- Strong practical experience with Python, SQL, APIs, automated testing, Git based delivery, CI/CD, and cloud services. Experience with Microsoft Azure, Databricks, MLflow, Vector Search, Azure OpenAI Service, or comparable platforms is strongly preferred.
- Ability to combine probabilistic AI techniques with deterministic rules, calculations, validations, and workflow controls to meet business, policy, audit, and operational requirements.
- Working knowledge of retrieval quality, prompt testing, confidence scoring, error analysis, model and prompt drift, guardrails, red teaming, privacy, security, and cost optimization.
- Experience collaborating with product owners, domain experts, full stack engineers, QA, security, data, and platform teams through design, UAT, production release, monitoring, and continuous improvement.
- Experience in financial services, mortgage lending, lending operations, servicing, broker channels, third party partnerships, or regulated technology environments is preferred.
Relationships:External: Engages implementation partners, cloud providers, data and AI platform providers, and specialist vendors on solution delivery, code and prompt quality, production readiness, and knowledge transfer to First National.
Internal: Works day to day within an AI Product or AI Platform Squad.
Working Environment and Physical Demands Analysis:- Office environment
- Periods of high volume with tight timelines
- Long periods of stationary position/sitting
- Prolonged periods of repetitive movement (i.e. using a keyboard and mouse)
- Long periods of time in viewing a computer screen
- Multi-tasking may include speaking to customers on a telephone call while looking up information on a computer program.
The team you'll join: Founded in 1988, First National is one of Canada's largest non-bank lenders. We provide residential mortgages exclusively through the mortgage broker channel and we are Canada's largest commercial mortgage lender.
First National has been consistently recognized as a great place to work and we are proud that our employee engagement feedback is higher than our industry partners.
We would like to thank all applications for their interest in this existing vacancy, but only candidates selected for an interview will be contacted. Artificial Intelligence is not used in our recruitment or hiring process for this role.#FNLOON 
