Small businesses power every economy. The founders behind them begin with big ambitions for where they want to take their companies—but most never reach that vision. Day-to-day, their time gets swallowed by operating the business instead of growing it: constantly switching between meetings, emails, and an average of 30 different SaaS tools, leaving barely 30 minutes for strategic thinking. Software was meant to simplify this. Instead, managing the chaos became a job of its own. We built Toyo first to solve this problem for ourselves—and now we’re making it available to everyone. 

Toyo gives founders and their teams a single platform where intelligent agents help operate the entire business: personal workflows, sales, marketing, operations, and finance. These agents understand your business context, run continuously in the background, and reach out through WhatsApp, SMS, iMessage, or voice when human input is required. 

We’re building the foundational infrastructure to make this possible: isolated compute environments, persistent storage, durable objects, and a secure credential layer that allows agents to take meaningful actions without introducing risk. OpenClaw demonstrated what’s possible when an agent has access to a full computer and long-term memory. We’re creating that same experience for business owners who need to trust an agent with their company—not developers experimenting on local machines. Security is embedded into the system architecture from day one. 
What We’re Building 

Small businesses drive the global economy, yet founders rarely get to focus on long-term growth. Their time is consumed by operational overhead—juggling tools, messages, and processes. Toyo is built to remove that friction. Our platform equips founders with AI agents that manage workflows across sales, marketing, operations, and finance—so leaders can spend more time building the business instead of running in circles. 

Our agents operate continuously and communicate across channels like WhatsApp, SMS, iMessage, and voice when human judgment is required. 

Behind the scenes, we’re developing the core primitives that make this safe and scalable: secure compute isolation, durable state, persistent memory, and credential systems that allow agents to take real-world actions responsibly. We run Toyo in production ourselves. A seven-person team operating with the output of a company three times our size. We’ve raised CAD $6M from Frontline Ventures and iNovia Capital, and our founders have previously built and exited companies acquired by Workday, MessageBird, and Fox. 

We’re looking for a Senior AI Engineer to help us scale the infrastructure that makes all of this possible.  

The Role 

You’ll join as our Founding AI Engineer, working closely with the founding team to design and evolve our agent infrastructure as we grow from early customers into a full platform. This is a senior individual contributor role with real ownership: you’ll write core systems, make architectural decisions, and tackle the hardest technical problems head-on. 

Your work will span two major areas: 

• Agent systems: session handling, agent communication, execution environments, and reliability.
  
  
• AI engineering practices: agents-in-the-loop for development, evaluation frameworks, and building reliable AI systems at scale.
  
  

The Stack
 
Our architecture has two tightly integrated layers:
 
Edge-native control plane:
Real-time APIs and session management run at the edge using durable, stateful primitives. We use Durable Objects for session registries, message streaming, live voice calls, and messaging bridges (SMS, WhatsApp, Slack). Long-running workflows survive failures, with a relational database behind a connection proxy for core data.

Secure agent VMs:
Each organization runs agents inside isolated virtual machines with full filesystem access, browser automation, and persistent state. Sessions connect via ACP (JSON-RPC over WebSocket). VM lifecycle management, snapshots, boot environments, and OS-level reliability are core parts of this role.

Application layer:
TypeScript across the stack. APIs built with Hono and end-to-end type safety. React 19 on the frontend. Enterprise SSO for authentication and provider-agnostic model routing so models can be swapped without changing agent logic.

What You’ll Work On
 

• Agent reliability: Designing human-in-the-loop checkpoints, approval workflows, and structured validation to prevent costly mistakes.
  
  
• Session infrastructure: Building resilient session lifecycles, multi-agent orchestration, and recovery after timeouts or failures.
  
  
• Secure VM environments: Hardening boot processes, snapshots, isolation, and OS-level recovery mechanisms.
  
  
• AI-native development: Defining how agents help build Toyo itself—issue triage, automated feedback loops, and evaluation pipelines.
  
  
• Event-driven agents: Scheduled agents, webhook-triggered workflows, and messaging integrations that keep working even when no one is online.
  
  

What We’re Looking For
 
You’ve fundamentally changed how you work with AI—not just adopted tools. You understand what works in production versus demos and experiment with new techniques as they emerge. Formal education matters less than what you’ve actually built.
 
Strong signals:
 

• Claude Code is your primary development environment
  
  
• You’ve shipped real-world agent systems
  
  
• Hands-on experience with LLM APIs, prompts, tools, structured outputs, and evaluations
  
  
• Deep understanding of edge-native, stateful distributed systems
  
  
• Experience working with VMs or containers and debugging OS-level failures
  
  
• Strong TypeScript skills and attention to correctness
  
  

Nice to have:
 

• Experience with agent protocols (ACP, MCP, etc.)
  
  
• WebSocket-heavy real-time systems
  
  
• Thoughtful approaches to agent security and credential isolation
  
  

Not a fit if:
 

• You require rigid specs and mature codebases to get started
  
  
• You view AI as a minor productivity boost rather than a fundamental shift in how software is built
  
  

What You Get
 

• Compensation: CAD $200K–$325K salary + equity
  
  
• Ownership: Direct influence on product and architecture from day one
  
  
• Culture: A small team that uses the same AI-first workflows we build
  
  
• Remote-first: No HQ, no commute, no bureaucracy
  
  
• Team retreat: Annual in-person offsite (last year: Sevilla)
  
  

How to Apply
 
Email hello@toyo.ai with the subject: Senior AI Engineer
Skip the cover letter. Send something you’ve built—a repo, write-up, or live project. Tell us what you find hardest about building reliable agent systems and how you’d approach it.