Struct

Struct built AI voice agents designed to handle phone-based tasks across multiple languages, targeting businesses that needed to automate customer service and outbound calling at scale. The problem was real: companies struggled with high labor costs for multilingual call centers and faced bottlenecks during peak demand periods. Customer service teams and sales departments experienced this most acutely, particularly in regions requiring non-English support. The pain was measurable—businesses could track call volumes, resolution times, and per-call costs. Yet alternatives already existed: traditional outsourcing firms, basic IVR systems, and emerging competitors like Retell AI and Bland AI were entering the same space. Struct's fundamental miscalculation lay in underestimating regulatory and compliance complexity. Phone-based AI agents operate in heavily regulated environments with strict consent, recording, and disclosure requirements that vary by jurisdiction. The warning signs were there: the company launched with "tens of thousands of calls daily" claims, suggesting they'd moved fast without fully addressing legal frameworks. This aggressive scaling without compliance infrastructure likely triggered regulatory pushback, making their business model unsustainable despite strong product-market signals.

Struct built multilingual AI voice agents handling tens of thousands of daily calls across multiple languages and geographies. Early behavioral signals appeared strong: customers actively deployed agents into production rather than running pilots, and call volume grew organically as existing clients expanded usage. The team measured genuine interest through actual call completion rates and customer retention metrics—not surveys or letters of intent. Initial traction showed real revenue from enterprises willing to pay for live agent deployment, suggesting authentic demand beyond stated interest. However, critical warning signs emerged. The market fragmented across languages and use cases, making unit economics difficult to optimize. Customer acquisition costs remained high relative to contract values, and churn accelerated as clients discovered integration complexity. The team likely underestimated how much customization each deployment required and overestimated how quickly the product could scale horizontally. By the time these structural problems became apparent, the company had already burned through runway without achieving sustainable growth metrics, leading to inactivity despite YC backing.