Senior Product Manager

Job description

*On-site | Vancouver Office – 675 W Hastings St.* #### Our Mission Our mission is to amplify human success with AI automation. In a dental lab, that means technicians spend less time on repetitive casework and more on the complex cases that need real skill, and labs take on volume they used to turn away. We automate the parts of the job nobody wants. We don't pretend that's the whole job. *Our vision is to become the operating system of the dental industry.* Not a tool a lab logs into. The system a lab runs on. #### **The Problem We're Solving** A dental lab loses hours every week to a question that shouldn't need asking: *where is that case?* Orders arrive from a dozen different manufacturer portals, and chasing them down has always been somebody's job. The Command Center ends that. It gives a lab full visibility of every case — what arrived, what stage it's at, what's stuck and why — updating continuously, in one place. Your job is to build the next generation of it around the way a lab actually works. ## #### **Why EviSmart** - 300 people. Two hubs: Vancouver HQ and Manila operations. - 145% year-over-year SaaS growth. - 28 countries. One platform. The dental industry's operating system. - An in-house AI model research and development team building proprietary intelligence. #### **Why This Role Exists** A lab isn't one job. It's designers, technicians, quality control, case managers, shipping, the owner — each with a different week and a different definition of what matters today. Full visibility means something different to each of them. The next generation of Command Centers has to be built around how those roles actually work. No written definition exists yet of what the next version of the Command Center must contain, what's optional, or what disqualifies one from shipping. You write that definition, and you write it from inside dental labs. Then you hold it, while everything else gets built against it. #### What You'll Own - **Define the product itself**: what each Command Center must contain, what is optional, and disqualifies one from shipping. Write the spec early and ship the first version. - **Own the moments the AI has to earn a user's trust.** Where automation acts on its own, where it asks first, where a person steps in, and what someone sees when the system isn't sure. - **Instrument every feature before it ships** — adoption funnels, cohort analysis, usage patterns. Baselines captured before launch, not after. The number this role is measured on is Weekly Active Role-Fit: the share of the lab roles a Command Center is built for who open it in a given week. The bar is 80% on pilots, and getting the tracking live is on you. - **Own the commercial case, not just the build.** Before anything ships: who pays, what the problem costs them, and what the value is worth in their terms. After it ships: whether they actually paid. You won't set the price, but you'll own the argument for what we're charging for, what's bundled, and what's worth paying more for. A product nobody pays for isn't a product. - **Drive user validation continuously:** conduct interviews, observe workflows, and build a user insight library the entire squad references for every product decision. Your first weeks include ten lab interviews and ride-alongs at three of them. The people you're building for are standing at a bench, not filling in a survey. - **Own the record of how decisions got made** — the problem statement, the acceptance criteria, and what was decided and why. We have two offices and memory is lossy. What you write outlives the conversation. ## #### What You'll Get - A seat where AI model decisions and product decisions meet. - Direct customer access across 2,000+ labs — the feedback loop is fast, the stakes are real, and user research here isn't a quarterly exercise. - A category-defining problem in a $40B industry where the incumbent software hasn't been rebuilt in a generation. - Production AI tooling from day one: Claude, Cursor, and LLM-powered workflows built in-house — not a sandbox, not a pilot. - Competitive compensation with salary range disclosed at offer. #### How We Work We ship before we're 100% certain. We write things down because we have two offices and memory is lossy. We debate loudly and move without resentment. We treat the customer's real problem as more important than an elegant internal process. If you've spent time waiting for permission to try something obvious — you'll notice the difference here immediately. #### #### What We're Looking For - **8+ years in product**, in B2B SaaS — workflow, automation, or platform. Not consumer, not content. - **You have been accountable for whether code ships.** You were an engineer before you were a PM, or you've had engineers report to you, or you still write code as part of your job. - **You've shipped an AI or ML product to production** and can describe the mechanics — how you chose a model, how you evaluated it, what you did about the cases it got wrong. - **You've owned a product surface end to end**, discovery through ship, with a named outcome. - **You've measured whether people kept using what you built** — retention, weekly actives, adoption by user type. Not launch counts. - **You've done customer discovery in person**, in the place the work happens. Apply today @ [EviSmart Careers](https://www.evismart.com/careers)