Deck provides an infrastructure layer enabling AI agents to interact with any software interface—web, native, or legacy—without requiring API integrations. The stack reveals a browser-automation-first architecture (Chromium, Puppeteer, Playwright, Selenium, V8) paired with GCP infrastructure and observability tooling (Prometheus, Grafana, OpenTelemetry), suggesting heavy engineering investment in reliability and scaling. Active projects around AI decision-making, self-diagnosis models, and bot-detection evasion indicate the core technical challenge: reliably automating interactions in high-security, login-protected environments at scale.
Deck was founded in 2024 and is based in San Francisco. The company builds infrastructure that allows AI agents to automate workflows across software interfaces without API dependencies, targeting scenarios where traditional integrations are unavailable or impractical. The product surfaces across web, native, and legacy systems. The engineering-forward hiring mix (split evenly between engineering and sales roles, with senior and manager-level leadership) reflects an early-stage infrastructure company balancing technical depth with go-to-market. Current hiring velocity is minimal, with active roles in the US and Canada.
Deck uses Chromium, Puppeteer, Playwright, and Selenium for browser automation; GCP (Cloud Functions, Pub/Sub, Firestore, BigQuery) for cloud infrastructure; PostgreSQL and SQL Server for data; and Prometheus, Grafana, OpenTelemetry for observability. Development languages include Python, C++, and Go.
Current projects include AI-powered browser automation, AI decision-making for autonomous navigation, self-diagnosis models for AI systems, and scaling the orchestration layer. The company is also establishing BDR processes and building an internal developer platform.
Other companies in the same industry, closest in size