The Opportunity Third Factor has spent 30 years helping leaders become better coaches and higher performers. We have trained over 40,000 leaders, we work with Queen’s University and the University of North Carolina, and we developed the 3x4 Coaching framework that organizations like RBC, BMO, Asana, Starbucks, and Deloitte use to help their leaders coach more effectively. Since 1991, we have helped over 100 athletes reach Olympic and Paralympic podiums.
The coaches we work with in sport provide inspiration for the work we do in corporate and academic contexts. Our work happens predominantly in classrooms and off-sites. What we cannot do is sit with a leader at 9pm the night before a tough conversation.
This is where ‘PJ: The Coach's Coach’ comes in. PJ extends what leaders learn from us into the moments classroom training cannot reach, giving them 1-on-1 real-time coaching support whenever they need it. PJ is live with paying enterprise customers, and ready for an engineer to own the technology, the architecture, and the roadmap.
You are not joining a 10-engineer team but you are not standing up an engineering org from scratch either. You are leading the next phase of a working product, with a team behind you: a development agency that knows the current codebase, a product manager and designer, and Carrie Drybrough (VP of Operations, your hiring manager). You have real autonomy - and also real support.
Beyond PJ You are Third Factor’s Lead Engineer, not just PJ’s. PJ is your biggest project coming out of the gate and where you will spend most of your first year but it is not the whole job. PJ is the first of several projects where we will use technology to change both what we sell and how we run.
Our operations team has mapped 35 internal workflows that are ripe for automation or rebuilding. Each of our three practice leads (coaching, collaboration, and resilience) has a roadmap with technology somewhere in it. As the most technical person in the company, you are the point of the wedge into that work: you help the business figure out what is worth building, then you build it, buy it, or direct the agency to build it.
Third
Factor is small enough that the most technical person in the building can help shape how the whole company works. About PJ
PJ runs on Gemini with a Node.js backend and a web frontend. It is built around the 3x4 Coaching framework: three Plays (Clarity, Competence, Recognition) executed through four Skills (Questioning, Listening, Feedback, Confronting). The interesting engineering problems are not the chat UI, but rather:
How do you route between fast and high-quality models without users feeling the seams?
How do you give PJ real memory of a coach’s people and conversations, not just a longer context window?
How do you measure whether the coaching advice is actually effective, not just plausible?
How do you meet enterprise security requirements (SOC 2) while shipping fast? Where the Product is Today The core coaching chat loop is live in production with paying enterprise customers A v1 of cross-conversation memory and retrieval just shipped. The next generation, with deeper coaching-specific context handling, is one of this role’s biggest opportunities
No automated AI evaluation pipeline yet. Building it, measuring PJ’s responses against the standard our practice has set, is a priority for this role
Agentic patterns (multi-step reasoning, tool use, orchestration) have not been built
SOC 2 compliance work is well underway What You Will Do Engineering & Architecture: Own the technical roadmap. Lead the build hands-on, with the agency working alongside you. Design infrastructure for real-time LLM inference, cross-conversation memory, and complex user state.
AI Engineering: Lead the LLM strategy. Prompt engineering, model selection, multi-model routing, and fine-tuning where it earns its keep. Build evaluation into the development cycle so we can measure PJ’s coaching responses against the standard our practice has set.
Memory & Retrieval: A v1 of cross-conversation memory just shipped. The next generation, designed around the unique demands of coaching, is the heart of your roadmap. We have early thinking on the shape this could take - we’re looking for someone to pressure-test that thinking, and bring the engineering judgment to execute.
Agency Partnership: Our development agency knows the current codebase best. Set technical direction and standards. Deploy their capacity against the work that benefits most from it.
Product and Business Partnerships: Work with coaches, facilitators, and business leaders who are exceptional at what they do but are not engineers. Explain trade-offs clearly. Listen for the domain expertise that should shape your architecture. Beyond PJ, help operations and our practice leads find where technology creates leverage, then make the call on what gets built or bought.
Operations & Security: Security, reliability, cost, and SOC 2 compliance live with you. Coaching conversations are private. They have to stay that way.
What You Bring
We are looking for a builder and technical leader, not a manager who codes occasionally. 5 to 8 years building production software, with real ownership of systems you shipped and kept running. You are Third Factor's only engineer, so beyond PJ you will carry other technical initiatives across the company, as we mature together. Must-Haves TypeScript and Node.js expertise. This is the stack. You should be productive immediately.
Shipped LLM-powered products to real users (1.5 to 2+ years). You know the difference between a demo and production.
Prompt engineering and context window management. Versioning, testing, managing system prompts across use cases. You understand how to structure and optimize what goes into the context window, not just what comes out.
Retrieval architecture. Embeddings, vector stores, chunking, retrieval-quality measurement. You have built RAG before. At PJ, a v1 is live; you will be taking it deeper around the unique demands of coaching memory.
Self-direction and plain communication. You set your own priorities and make architectural decisions without a senior engineer above you. You explain trade-offs to non-technical stakeholders without condescension, and listen for the parts of the domain you do not yet understand. You keep several initiatives moving at once, and you know which one needs you right now.
Vendor or agency management expertise. Nice-to-Haves
Real interest in coaching, leadership development, and how adults learn
LLM evaluation pipelines (RAGAS, DeepEval, or your own).
Building systematic evaluation for PJ is a priority;
experience here gives you a head start.
SOC 2 or similar compliance framework experience
Agentic AI patterns (multi-step reasoning, tool use, orchestration). We have not built these in PJ yet, but we know we need to.
Python proficiency for AI tooling and evaluation work
Founding or sole-engineer experience at a product company The In-House Stack Aside from the PJ tech stack, Third Factor uses several tools in-house:
You own or co-own: GitHub, LLM APIs, Microsoft 365, Vercel A real plus: integration and no-code tooling — Zapier, Knack, and similar
Familiarity welcome, all learnable: ActiveCampaign, Adobe Creative Cloud, Dropbox, Granola, Gravity Forms, Mailchimp, Mentimeter, Salesforce, TalentLMS, Typeform, WordPress, and Zoom What We Offer Competitive Compensation: $140,000 - $180,000 CAD, depending on experience
Full health and dental benefits
Remote-first: This is a remote role and you will work from your home office. We will consider candidates from anywhere in North America, but have a preference for Greater Toronto Area-based candidates, where we gather once per quarter for all-hands work-together days.
Career development budget: AI conferences, courses, etc.
Direct access to leadership: You report to our VP of Operations, work daily with the product manager, and sit close to the CEO and the practice leads.
Impact that compounds: Every feature you ship affects how thousands of leaders develop their people.
Why This Role The AI coaching space is wide open. The company that builds genuine long-term memory and contextual understanding into coaching AI, not just chat history, will define the category. This is a rare intersection. A greenfield AI product opportunity, at a company with 30 years of proven methodology, paying enterprise customers, and product-market fit. You get startup-level autonomy and impact without startup-level uncertainty.
Our Hiring Process
This is a remote role, but we always include an in-person stage when hiring full-time people. The process comprises five stages, and we will move quickly: Application. Everything we need from you is at the bottom of this posting.
Intro call, about 30 minutes.
Mutual fit: how you work, what you are looking for, and whether this is a role that excites you.
Technical deep dive. A bounded design brief, roughly two to three hours, on a real PJ problem. We review it with you live. Not a multi-day build.
Interview Day. In-person day in Toronto. Finalists spend a day with our leadership team and practice leads at our Toronto office. (We will reimburse mutually agreed flight, hotel accommodation, and other travel costs). Final conversation, references, background check, and offer.
How To Apply Send the following to
[email protected]: Your resume A link to your GitHub, portfolio, or anything else that shows your work. A sparse public GitHub is not a strike against you; we read code together at the deep dive.
Short answers (no more than ~250 words, each) to these two questions:
Tell us about an LLM feature you shipped to real users. What did v1 look like, and what did v2 look like after real usage? How did it change, and why?
Name one thing you believe about building LLM products that most people using and/or building them get wrong. Why is that? Only complete applications, resume plus link plus both answers, will be reviewed. Please apply by email rather than LinkedIn Quick Apply.
📌 Lead Engineer (Toronto)
🏢 Third Factor
📍 Toronto