Senior Software Engineer - Applied AI

Description

The company is an investment research and data firm. It provides independent research, built with technology and design, to investors worldwide, from individuals to professionals who manage money for millions.

This is a senior applied engineering role building the systems that bring AI into production: data pipelines, LLM integrations, agentic workflows, tool interfaces, and evaluation frameworks. Success is measured by shipped, maintainable systems rather than research. A good fit likes working in a field that changes quickly and builds architectures with swappable parts, so new models and techniques are adopted based on evidence. The team covers cloud architecture, local model inference, evaluation design, agent orchestration, API and tool design, and AI-assisted data enrichment, and it needs engineers who can set up trustworthy sources of truth, catch regressions, and know when a deterministic approach beats an AI one.

Responsibilities

• Connect to hosted LLM inference through the central model gateway for extraction, classification, and agent orchestration, and compare open-weight models with hosted options on cost and performance.
• Build and maintain adapters that sync source systems, such as the CMS, event platforms, editorial tools, and research libraries, into a governed dataset while leaving source-of-record logic where it lives.
• Create and run an eval harness: maintain golden questions, catch regressions before release, and make eval results the gate for shipping.
• Design MCP tools and API or GraphQL interfaces that keep 'fuzzy' retrieval (ranked candidates with confidence scores) separate from 'exact' governed lookups that are deterministic and carry provenance.
• Extend the pipeline that validates, resolves, and versions governed entities into a serving layer.
• Build and run LLM extraction and classification agents that propose structured data for human review rather than publishing AI output without review.
• Own structured content modeling in the headless CMS, including schema and versioning decisions that other teams depend on.
• Add telemetry to pipelines and served surfaces for freshness, adoption, and answer quality.
• Join and help run the weekly eval review and the biweekly skill-library session, bringing reusable agent tooling to the team.
• Mentor engineers moving into applied AI work, especially on eval design and agentic coding practices.
• Feed governance and vocabulary decisions into org-wide standards where relevant, instead of duplicating them.

Requirements

• 5+ years of software engineering experience, including designing production APIs and data pipelines.
• Use agentic coding tools every day as a core part of your workflow.
• Hands-on production experience with LLM integration: Skills, MCP, RAG, structured outputs, and tool or function calling.
• Strong Python for eval tooling, service code, and API development with FastAPI or similar, plus working TypeScript/Node.js for application integrations.
• Experience running production services on AWS or an equivalent cloud, including containerized workloads and infrastructure as code.
• Judgment on security and privacy in AI systems: handling sensitive data properly and guarding against prompt injection, data leakage through prompts, and unsafe or unattributed model output.
• Experience meeting latency, reliability, and cost targets in production, including managing token costs for inference.
• Evaluation literacy: you can explain an eval you built, what it caught, and how you handled canonical truth, variance, and regression cases.
• Experience with REST or GraphQL API design.
• Working knowledge of data pipeline patterns: idempotency, versioning, and staged architectures.
• Strong written communication: you can write a clear design doc, explain a tradeoff to a non-engineer, and document decisions others will build on.
• Comfortable with ambiguity and solving problems creatively, with a bias toward shipping over ceremony.
• Experience evaluating or benchmarking open-weight models for cost and performance at inference time.
• Experience with cloud-hosted inference services such as Amazon Bedrock or Azure AI/Cognitive Services, and with central LLM gateways such as LiteLLM.
• Experience with headless CMS platforms and structured content modeling.
• Experience in a regulated or compliance-sensitive domain where provenance and auditability matter.
• Formal model training or ML research credentials, such as pretraining or fine-tuning research, are not required. This is an applied systems role, and the company is looking for builders rather than researchers.

Conditions

Hybrid role based in the Toronto office, with at least 4 days onsite each week.

In most company locations, the hybrid model is four days in the office each week, so teams can work together regularly.

Total cash compensation (base plus bonus target) is 101,800.00 to 207,000.00 CAD.

If you accept an offer, you must disclose your personal and related investments to the Compliance team, on a timeline that varies by region. Compliance reviews them against the Code of Ethics, and any conflicts of interest must be liquidated immediately. Depending on department and location, some accounts must be held with an approved broker, for example all U.S. employee accounts. If that applies and your account is elsewhere, you must move your holdings to an approved broker.

Benefits

• A range of other benefits is available to add flexibility as needs change.
• You get tools and resources to work with global colleagues, wherever you are based.