Toronto (Financial District) · In office, five days a week · Full-time
The Product Builder owns a product problem from the first client conversation to working software in the client's hands. You find the problem worth solving, decide what to build, and build it with coding agents.
The agents write most of the code. You own the rest: the problem, the decisions, and whether the result is right.
About 15Rock
15Rock pairs top experts with custom AI agents and workflows to deliver institutional-grade research and analysis.
Our clients include large asset owners and some of the largest advisory firms in North America. We also work with fast-growing, complex companies on value creation, corporate strategy and M&A, and we help clients expand into new markets and industries.
Our products include TrendCap, our research and analysis platform, and Novilo, our platform for the full bid lifecycle.
Our work in numbers
Our agents and experts have analyzed about 3,000 companies and institutions across 23 industries.
- 2,000+ companies rebuilt in financial models, from P&Ls and margins to working capital and unit costs
- 4,500+ company-level evaluations across delivered strategies and models
- US$650B in combined annual revenue across our deep-dive companies
- US$20B in enterprise value identified in our value creation plans
- US$5.6B in new revenue opportunities sized
- US$4B+ in cash tied up in working capital, found in accounting footnotes
- US$3.2B in annual EBITDA improvement sized across 46 operating companies
- 1,000+ firms mapped to 170+ defence supply chains, including suppliers that industry codes miss
Some of what we have found
Company details are left out or rounded to protect confidentiality.
- On a drone assembly line, parts spent 99.8% of their lead time waiting. Our redesigned layout is modeled to roughly triple throughput.
- We rebuilt a large North American telecom carrier's management structure from public data and sized more than US$400M in structural savings.
- On a long-term naval support contract, more than C$100M in performance fees depended on fleet readiness. We traced the shortfall to slow parts-planning data, not a lack of technicians.
- For a luxury carmaker, our model showed a planned flagship store would lose money every year on new-car margins alone. A seasonal location cut modeled occupancy costs by about 90%.
- Across 66 Canadian hospital networks, we benchmarked 100 indicators per weighted patient case and sized about C$180M in savings from clinical supplies and surgical scheduling.
What you will do
- Meet clients and users. Learn how they work today and where it breaks down.
- Decide what to build and what to leave out. Write a short brief that people and agents can build from without guessing.
- Build it end to end with coding agents: the interface, the API, the data and the AI workflow behind it.
- Check the work before a client sees it. Read the code and test the outputs against the sources.
- Ship it, watch how people use it, and improve it.
- Turn what works into workflows we can reuse across clients and industries.
Problems you could take on
Each one comes from our current client work and roadmap.
- From a study to a product. We map supply chains and value chains for clients, often hundreds of companies at a time. Make that a product clients can run and question on their own.
- Always on. Most research is a snapshot. Build the layer that watches a company, a portfolio or a market and says what changed and why it matters.
- Price with evidence, or abstain. Bid teams want to know what a contract is likely to be awarded at. Build tools that find comparable awards and give a price, or say plainly when the evidence is too thin.
- Should-cost from the outside. Estimate what a product should cost to make, using public data and supplier evidence, with every assumption visible.
What we are looking for
- You have shipped software that people use, and you can show it to us.
- You use coding agents such as Claude Code or Cursor every day. You know where they go wrong, and you catch it.
- You work across the stack: interface, back end and data.
- You have taken LLM features past the demo stage, with evaluations that tell you when the output is wrong.
- You have product judgment. You find the real problem and cut scope without being told to.
- You write clearly. Your briefs and pull requests are short and easy to act on.
- You can walk a senior client through your work and answer hard questions about it.
Useful, not required
- Experience in finance, consulting, procurement or public-sector bids
- Comfort reading financial statements and financial models
- Data pipelines that pull from public sources such as company filings and procurement portals
- Enough design sense to ship a clean interface without a designer
What this role is not
- A product manager seat that writes specs for others to build
- An engineering seat that waits for tickets
- A prototype seat. What you build goes to clients.
Two things to know before you apply
This is an in-person job, five days a week in our Toronto office. Decisions here happen fast and face to face, and you will be part of them.
The work is checked closely. Clients, Partners, investors and executives use what you build to make decisions. When something is wrong, there is massive impact.
Compensation and benefits
- Base salary
- Equity under the company's equity incentive plan
- Health and dental coverage
- RRSP matching
- AI tools without limits: frontier models and coding agents, no seat caps, no token budget to manage
- A small team, so you are in the room for product and architecture decisions, and you hear what clients say about what you built
Applying
Apply through LinkedIn with your resume and a link to something you built and shipped. A short demo video is welcome. Then answer two questions in a few sentences each:
- On that project, what did you decide not to build, and why?
- Describe a time an AI tool gave you code that looked right and wasn't. How did you catch it?
No cover letter. You must be legally entitled to work in Canada.
This posting is for an existing vacancy. 15Rock uses AI-assisted tools to help organize, screen, and assess applications. People review the evidence and make all interview and hiring decisions.
15Rock welcomes applicants from different backgrounds and learning paths. If you need an accommodation during the hiring process, please tell us what would help.
