Director of Software Engineering (AI Workflows & Ecosystem)
Canada {{REMOTE}}
Senior Level
Full-Time
About the role
- Jobber has AI in production, but not yet at its full potential
- We already have AI answering calls, drafting responses, and powering parts of our product. But today, those systems are still fragmented. Some teams are ahead. Others aren’t. Some workflows are intelligent. Others are still manual. And most importantly, the system doesn’t yet think across the product
- A service pro still has to:
- Manually follow up on jobs
- Piece together context across workflows
- Decide what to do next
- The platform doesn’t proactively help them run their business. That’s the gap
- The opportunity is to evolve Jobber from: AI-powered features → AI-powered workflows → AI-powered business operations
- This role owns that shift. Not a team. Not a feature. The system
- You’re building for people who don’t have time to think about software
- A plumber finishing their last job at 6 pm
- A cleaner managing 30 clients and 5 employees
- A landscaper juggling scheduling, payments, and follow-ups
- They’re not asking for “AI.” They’re asking:
- “What should I do next?”
- “Why didn’t this job convert?”
- “Who should I follow up with today?”
- And eventually:
- They shouldn’t have to ask at all
- The Director who succeeds here will understand:
- This isn’t about building clever systems; it’s about building systems that remove thinking from already overwhelmed people
- End-to-end ownership of Jobber’s AI system layer. You’re not owning a single team. You’re owning how intelligence flows across the entire product
- Product + Platform Scope
- AI Foundations (models, orchestration, evals, guardrails)
- Copilot (user-facing intelligence layer)
- Automations (workflow execution layer)
- Platform Experience / Marketplace (integration + ecosystem surface)
- Emerging surfaces (voice, messaging, cross-product intelligence)
- You are responsible for:
- How decisions get made inside the system
- How context moves across workflows
- How actions get triggered (and when they shouldn’t)
- How we evaluate whether AI is actually working
- This includes:
- Agentic workflows (reason → decide → act → evaluate)
- Cross-product context (jobs, customers, payments, communication)
- Reliability, safety, and failure modes
- Developer experience for building on top of AI systems
- Team Structure
- ~30 engineers across 4–6 teams
- 4–6 EMs / Sr EMs reporting into you
- Close partnership with Product, Design, Data
- WHAT “GOOD” LOOKS LIKE
- Not “we shipped AI features.”
- Instead:
- The system proactively recommends and takes actions
- Teams build on shared AI primitives, not reinventing them
- AI output is reliable, measurable, and improving over time
- Engineers trust the system, and move faster because of it
- Customers feel like the product is working for them, not just responding
- Define how AI should work across Jobber, not just within a team
- Build and evolve a multi-team org to execute on that vision
- Make tradeoffs between speed, quality, and safety
- Push teams beyond feature thinking into system thinking
- Challenge assumptions, including leadership’s
- Drive adoption across engineering, product, and the company- We are looking for someone who has:
- Observability and failure handling
- Guardrails and safety in real systems
- Evaluation (offline + online)
- Agent orchestration (not just prompts)
- Built real systems where AI makes decisions and takes actions in production
- That means experience with:
- Tool use and workflow execution
- You don’t need to code daily, but you must be able to reason at the system level
- Tradeoffs between autonomy vs. control
- Built organizations that scale (not just teams that ship)
- You’ve led orgs through complexity, not just growth
- Managed managers across multiple teams
- Driven cross-org alignment in ambiguous spaces
- You understand how user workflows connect end-to-end
- You care about customer outcomes, not just technical output
- You’ve partnered deeply with Product and Design
- You think in systems, not features
- You’ve built or led production LLM/agentic systems
- You have opinions about evaluation, reliability, and safety
- You understand what actually works (and what doesn’t)
- You’ve seen systems fail and improved them
- You know when to iterate and when to redesign
- You can move fast without breaking everything
- You’ve balanced shipping vs infrastructure vs tech debt