Fullstack Software Engineer (In-person, Toronto, Canada)

Provision makes collaborative document management software for the construction industry. Instead of manually reading through thousands of pages of documents and revisions, Provision organizes and extracts information so constructors can save hundreds of hours per bid and make more money.

Provision is automating construction. We help contractors speed up their work, reduce risk, and eliminate errors that cost the industry more than $500B every year. We are building the most ambitious vision of what is possible with AI in construction, and we expect our engineers to be AI-forward in how they build.

• Build and evolve full-stack product surfaces, backend services, and document-processing systems.
• Develop prototypes that prove key product concepts and quantify technical constraints for advanced AI pipelines.
• Integrate cutting-edge AI into customer-facing automation workflows alongside machine learning engineers.
• Improve reliability, performance, and quality across web applications, backend services, and AI-assisted workflows.
• Collaborate with product, design, customers, and internal users to turn ambiguous construction workflows into useful software.

• 3+ years of software development experience, or 5+ years of professional experience building software in lieu of a degree.
• Production full-stack web application experience, especially with React, Python, and Django.
• Strong software design, debugging, performance optimization, and test automation skills.
• Practical proficiency with AI coding assistants, LLM-based development workflows, or AI-native tools.
• Sound engineering judgment: you use AI to raise throughput without outsourcing taste, validation, security, or maintainability.

• You move quickly, take ownership, and communicate clearly in ambiguous environments.
• You care about users and connect technical choices to customer value.
• You hold a high bar for engineering quality, security, and design.
• You believe proficiency with AI is now part of the craft of effective software engineering.

Our technical challenge is taking hundreds of different document formats, processing them automatically to create a coherent data representation and then building an application to make it easy to understand for construction users.

Our ML Document Processing pipeline is written in Python and we use pytorch, spacy, nltk, and pandas where appropriate. We primarily write code in Python, using the Django web framework and in Typescript for a react web application. We host on a hybrid cloud setup across GCP and AWS. We host all of our application data in Postgres.

Our interview process is designed to be practical, transparent, and representative of the work you would do at Provision.

1. Introductory call — 20 minutes

A short conversation about Provision, the Full Stack Software Engineer role, your experience, and what you are looking for next. 2. Online technical interview — 2 hours

A practical technical session covering full-stack development, product judgment, debugging, testing, system design, and how you use AI tools in your engineering workflow. 3. In-office interview day

Visit our Toronto office, meet the team, learn how we work, and get a feel for the environment. The day includes a collaborative whiteboarding session based on a realistic product-engineering problem. 4. References and offer

We will complete focused reference conversations, answer any remaining questions, and move quickly to a final decision and offer.

We will explain the format and expectations before every stage and aim to communicate next steps promptly.