Role: Senior AI Engineer
Toronto, ON
Work Schedule: Hybrid, Tuesday to Thursday, 8:30 AM to 5:00 PM EST (3 days per week required in office)
Mandatory Skills:
- JVM Engineering (Java/Kotlin)
- Knowledge Graphs (Neo4j, Cypher, GraphRAG)
- MCP (Model Context Protocol)
- Production Platform Engineering (CI/CD, Docker, Kubernetes, Observability)
- Document Generation & Rendering (Apache POI, PDFBox, pptxgenjs, Office Open XML, Word/PPT/PDF generation, template management)
Required Qualifications
These apply to everyone we will consider.
- 6-10 years building production software, including recent hands-on delivery of AI or large-language-model systems beyond prototypes.
- Strong JVM engineering (Java; Kotlin or Scala a plus) with solid practices: Git workflows, code review, automated testing, structured logging, and clean design.
- Hands-on experience with modern GenAI patterns: prompt engineering, structured or JSON outputs, tool and function calling, retrieval-augmented generation, and agentic workflows.
- Experience designing and querying a graph or document database (Neo4j and Cypher, or MongoDB Atlas) and using vector search and embeddings for semantic retrieval.
- A track record of owning your own delivery path: you have taken something you built through a pipeline into production and operated it, rather than handing it over.
- Practical experience evaluating non-deterministic systems: test design, quality scoring, regression suites, and translating evaluation into business-ready acceptance criteria.
- Demonstrated ability to design and explain solution architecture (data flow, runtime flow, interfaces, failure modes, and controls) and to explain model behaviour, limitations, and trade-offs in plain language.
And real strength in one of these two adjacent areas
- Platform and reliability: a major cloud (Azure preferred), containerized deployment with Docker and Kubernetes, CI/CD, observability, and automated quality gates on a service you ran in production.
- Deliverable generation and rendering: producing Word, PowerPoint, PDF, or Excel output programmatically with libraries such as Apache POI, PDFBox, or pptxgenjs, making that output deterministic and testable, and moving comfortably between a JVM service and a Node.js rendering toolchain.
Preferred Qualifications
- Experience with the Model Context Protocol (MCP), building tool or resource servers and clients, and with agent-to-agent (A2A) interoperability.
- Experience integrating with low-code agent platforms such as Microsoft Copilot Studio.
- Experience with event-sourced or workflow frameworks (for example, the Akka SDK, Temporal, or similar) for long-running, restart-safe processes.
- Experience with cloud AI services (for example, Azure OpenAI or Azure AI, or equivalent), GraphRAG, and document-intelligence or OCR pipelines.
- Experience building conformance, golden-output, or contract-test harnesses.
- Comfort across languages: Python and Bash for tooling, and the ability to read a Node.js codebase as readily as a JVM one.
- Familiarity with Office Open XML internals, or with rendering diagrams and charts programmatically (SVG, layout engines such as elkjs, or headless rendering).
- Experience implementing GenAI guardrails and delivering under formal AI or model-risk governance.
