Real-world sandboxes to test agents and agent-facing software
Developers lack safe, scalable ways to test agents and agent-facing software against realistic external-service behaviors and edge cases before production.
A sandbox platform that runs agents/code against API-compatible replicas of services like Stripe, Slack, and Google Drive, enabling parallel environments, natural-language seeding, and workflow observability.
Developers and engineering teams building AI agents or integrations that interact with third-party SaaS and APIs.
Co-founder & CEO @ Arga Labs. Built an internal dev tool @ Amazon to automate complex engineering workflows, resulted in 10+ recurring weeks/year of engineer hours saved across multiple teams, and prevented VP escalations due to bug issues. Before that, I was a fencer on the Canadian Junior National Team and researched cognitive systems at UBC
Co-founder, CTO @ Arga Labs Prev. SDE @ Stripe, quant @ GS Skipped high school, graduated @ 19 Pro player for Identity V, quit because I had to finish my finals :(


