Applied AI Engineer

Position overview

AI features and model selection

• Build production features using Claude, Gemini, OpenAI, AWS Textract, LandingAI, and other suitable services to extract data from technical documents and drawings.
• Benchmark vendor and open-source options for accuracy, latency, and cost; own routing and fallback logic.
• Create schema-constrained extraction, validation, and confidence scoring that sends uncertain results to human review.
• Develop RAG pipelines and vector indexes for product matching and specification cross-referencing.
• Deploy and operate self-hosted models where document volume, latency, or data sensitivity warrants.
• Build clean, tested Node.js or Python microservices for core applications.

Evaluation, MLOps, and data

• Maintain golden datasets and automated evaluation suites; run regression checks before prompt, model, or retrieval changes ship.
• Set accuracy and retrieval-quality targets, report results, monitor output quality, and detect drift and common failures.
• Own confidence thresholds for automated processing versus human review.
• Integrate AI services into AWS cloud microservices and help establish CI/CD, testing, and evaluation before deployment.
• Monitor accuracy, inference speed, latency, API failures, and token costs per document.
• Clean, process, and label technical data; build data-preparation pipelines and maintain labeled corpora for evaluation.

Collaboration

• Partner with Product Managers and Software Engineers to turn specifications into AI features and set realistic accuracy expectations.
• Participate in code reviews, technical spike evaluations, and sprint planning to support timely delivery.

Required qualifications

• 3+ years of professional software development experience, with at least 1–2 years building and deploying practical AI solutions.
• Advanced proficiency in Node.js/TypeScript or Python; strong candidates in one are expected to become productive in the other.
• Hands-on experience with LLM applications, prompt engineering, structured outputs, and vector databases.
• Familiarity with LangChain/LlamaIndex or a clear perspective on when to skip them.
• Experience in unstructured data extraction using PDF parsing, OCR, or multimodal LLMs.
• Demonstrated approach to measuring AI quality with golden datasets, regression testing, and accuracy measurement.
• Experience with RESTful APIs, relational databases, Git, Docker, CI/CD, and basic unit and integration testing.

Preferred qualifications

• A relevant bachelor’s degree is nice to have.
• Open-source model or embedding-model fine-tuning, PyTorch, and production operation of self-hosted models are nice to have.
• AEC, PropTech, CPQ, e-commerce, or B2B SaaS experience is nice to have.
• AWS or Azure services such as SageMaker, Lambda, S3, or ECS are nice to have.

Compensation

• Base compensation is $100,000–$140,000 CAD per year, plus benefits.

Workplace

• This is an in-office position in Scarborough, Ontario.

Benefits

• Benefits for eligible full-time employees include medical, vision, and dental coverage, disability and life insurance, paid time off, and RRSP match.
• Other listed benefits include casual dress, company events, food discounts, an on-site gym, parking, and career advancement opportunities.

Application

• The posting requests a resume and salary expectations by email; it also lists questions about production AI, document processing, and AI evaluation experience.