By Soumoshree Mukherjee
Simplex officially launched its production-grade web agent platform, setting out to solve one of the most persistent problems in enterprise AI — automating workflows across legacy portals.
Founded by MIT graduates Shreya Kapoor and Marco Nocito, the startup emerges from stealth with a developer-first solution built to integrate seamlessly with modern AI stacks while reliably handling outdated, complex web systems.
Simplex’s origin story begins with a fundamental challenge in AI engineering, legacy systems are everywhere, and they’re not built for modern automation. As companies adopt AI agents to handle high-volume, repetitive workflows, the lack of integration with these portals becomes a bottleneck. Existing tools often fail at critical tasks from invoice edits to form submissions, because they lack the control and transparency required for production-grade deployments.
READ: Regulating AI: Sanjay Puri on policy, challenges, and ethical innovation (November 1, 2024)
Kapoor, who holds a master’s degree in artificial intelligence from MIT, wrote in a LinkedIn post: “We just wanted to build. Turns out, that’s exactly what startups are about too! Build product, build relationships, make your users happy.”
From AI uncertainty to operational reliability
Where current argentic solutions fall short often getting stuck on broken buttons, hallucinating actions, or mistyping shipping details, Simplex steps in with a powerful alternative. The platform equips developers with both agentic and non-agentic tools to precisely define workflows, eliminating ambiguity and failure during execution. According to the company’s public statement, Simplex’s custom web agents are designed with features like real-time error handling, session replays, detailed logs, and automatic CAPTCHA and two-factor authentication (2FA) handling.
Customers are already using Simplex in production environments for tasks such as submitting outbound call data into freight management software, editing customer invoices, and tapping into internal website APIs. Its infrastructure supports hundreds of concurrent browser sessions, which makes it highly scalable for growing operations.
Kapoor and Nocito bring deep technical expertise to Simplex. Prior to founding the company, both worked on computer vision systems at Tesla and Waymo, and within advanced MIT research labs. Their combined experience in building reliable, scalable automation systems translates directly into Simplex’s product ethos: control, observability, and safety in critical workflows.
The team’s academic and professional pedigree, combined with their real-world insight into the pitfalls of web automation, fuels a platform that’s as robust as it is intuitive. As the AI tooling landscape continues to expand, Simplex is carving out a vital space for developers who demand precision and scale in production workflows.
Simplex is now part of Y Combinator’s Summer 2024 cohort and is already attracting attention from engineering teams looking to eliminate repetitive manual tasks at scale. Whether it’s aggregating financial data or filling out forms on outdated portals, Simplex offers a streamlined, dependable alternative to building one-off, fragile bots or begging vendors for APIs.
As Kapoor and Nocito put it, they’re not just building another automation tool, they’re “engineering web agents into the core infrastructure of future AI workflows.” With Simplex, developers no longer having to choose between speed and reliability, they chose to build a tool with both.
For those interested, a live demo and interactive playground are available on Simplex’s official website, and the team is currently inviting API access requests from developers and enterprise partners.
