Why a local-first, human-layer DLP had to exist

I'm Shubham Gupta, a Perth-based engineer and automation specialist. Building serverless applications, enterprise integrations, and AI-driven workflows at companies like Woodside Energy, I saw firsthand how easily sensitive data slips through the cracks — not through some sophisticated attack, but through an ordinary person uploading the wrong file to the wrong place.

The tools meant to stop this were built for a different era. Legacy DLP targets managed endpoints and technical staff, ships file content to a vendor cloud to inspect it, and takes months and a network project to deploy. Meanwhile the real exfiltration path moved to the browser — cloud storage, webmail, Slack, and, increasingly, AI tools that everyone in the company now uses daily.

Watch Dog exists to close that gap the right way. It intercepts uploads at the browser layer and scans them locally — so the content never leaves the device — then gives the organization a content-free audit trail and policy engine. It protects the 80% of staff heavy DLP ignores, and it deploys in an afternoon.

My approach to software has always been the same: lean, intuitive tools that deliver real value without a manual. Watch Dog is that philosophy applied to data loss prevention — powerful for security teams, invisible to everyone else. We're early, we're honest about where we are, and we'd love to build it with design partners who feel this problem.