Provision is a two-year-old startup applying machine-learning pipelines (PyTorch, TensorFlow, RAG, scikit-learn) to construction document analysis. The tech stack reveals a dual focus: ML infrastructure for document understanding and a sales/marketing layer (HubSpot, Salesforce, LinkedIn Ads, Google Ads) backing early customer acquisition. Active projects span automated risk detection, claims surfacing, and dispute prevention — all addressing construction's contractual friction point.
Provision helps construction professionals identify and mitigate risks embedded in project documents. The product sits at project intake, where teams upload contracts, specifications, and related paperwork; the platform surfaces hidden liabilities and workflow blockers. Based in Toronto with a 9-person team (heavy on engineering and leadership), Provision is building AI document understanding as the core, supported by demand-generation and adoption tooling. Pain points tracked include workflow automation, error elimination, and integration adoption — all typical signals of early construction software selling into process-heavy buyer workflows.
Python, PyTorch, TensorFlow, scikit-learn, RAG, HubSpot, Salesforce, PostHog, and Procore. The ML stack suggests document-understanding focus; sales tooling indicates B2B motion.
Toronto, Ontario, Canada. The company was founded in 2022 and is privately held with 2–10 employees.
Other companies in the same industry, closest in size