Who you are
- You're a builder first — you've seen that the highest-leverage engineering right now multiplies everyone else, and you want to do it somewhere with real ownership and zero red tape
- 7+ years of software engineering, including internal platforms or tools that measurably increased team output
- 1–2+ years building LLM-powered systems in production — real users, real reliability requirements, not prototypes
- Strong production Python and/or TypeScript, plus the full-stack range to ship a usable interface when a workflow needs one
- Deep, practical knowledge of agentic systems: tool use / function calling, orchestration, context engineering, structured outputs, memory, and background-agent patterns
- An eval-first mindset — success metrics before you build; observability, guardrails, and cost controls as part of the system
- Integration and auth chops across an enterprise SaaS stack: REST, webhooks, event-driven patterns, OAuth / OIDC / SAML
- Security and governance judgment — scoped, auditable access, human-in-the-loop design, and a transparent sense of what an agent should never do in a financial company
- Workflow-discovery skills: sit with a non-technical team and turn how they work into an automation spec with measurable targets
- Cloud fluency (AWS) and comfort with CI/CD
- Experience with agent frameworks and SDKs (Claude Agent SDK, OpenAI Agents SDK, LangGraph, Pydantic AI), and with MCP server design specifically
- Durable-execution or workflow-orchestration experience (e.g., Temporal) for reliable long-running agents
- Corporate-engineering / IT-systems depth: identity providers (Okta, Google Workspace admin), ITSM and approval workflows, or SaaS administration at scale
- An iPaaS / automation-platform background (Workato, n8n, Zapier, Retool), with transparent opinions on when no-code is the right answer
- Fintech or other regulated-industry experience, and a working sense of what SOC 2 / ISO 27001 mean for AI systems
- Power-user habits with AI coding tools (Claude Code, Codex, Copilot) — you'll be building the environment that makes everyone else one too
What the job involves
- We're hiring a Senior Corporate Engineering & AI Systems Engineer to build the internal platform that lets every Floater scale themselves — so each person feels like a team of ten
- You'll own three things: the harness (the secure, governed layer connecting AI models to Float's internal systems, with identity-aware access, tool connectors, guardrails, and audit trails), the platform (the paved road that lets any Floater compose, run, and share automations), and the background agents (always-on workers that triage queues, reconcile data, and draft responses while Floaters sleep, escalating to a human when judgment is needed)
- This is not an evaluate-some-tools-and-write-a-policy role — you'll be in the room when a team describes a problem, and shipping the fix, often the same week
- Build the harness: authenticated, scoped, auditable connectors and MCP servers linking LLMs to Slack, Google Workspace, Salesforce, NetSuite, Zendesk, and Float's own product
- Ship background agents that run multi-step workflows end to end — with the tool routing, memory, retries, human-in-the-loop approvals, and audit trails that make them trustworthy in fintech
- Build the skills-sharing layer that turns one team's automation into everyone's capability, and drive the adoption that makes it worth building
- Embed with teams to map how they work and turn ambiguous problems into shipped systems — defining evals before you build and instrumenting reliability and cost as you go
- Own governance with Security and Risk: non-human identity, scoped permissions, prompt-injection defense, and privacy guardrails that enable rather than block
Benefits
- Competitive coverage of medical, dental and vision insurance for employees
- Education & learning stipend for personal growth and development
- Flexible vacation time
- Work from home stipend to help you succeed in a remote environment
📌 Senior Corporate Engineering & AI Systems Engineer (Toronto)
🏢 Float
📍 Toronto