Traba

Traba built Scout, an AI-powered interview system, to solve hiring bottlenecks in industrial supply chains. Scout verifies workers in the logistics and manufacturing fields and matches them to suitable worksites.

The initial Scout V1 used a single-agent structure and supported only English, but it evolved into a system capable of large-scale operation after adopting ElevenLabs's real-time TTS and STT technology.

The key upgrades are as follows:

• Multilingual support: Expanded the range of workers that can be supported through real-time switching between English and Spanish.
• Multi-agent orchestration: Specialized agents for onboarding, verification, logistics, and FAQ divide up and handle the conversation.
• Duplicate logic removal: A pipeline that removes duplicate questions reduced redundancy per candidate by up to 20%.
• Custom evaluation framework: Provides the ability to evaluate answers according to each client's unique criteria.

Scout currently handles more than 50,000 interviews per month, automating 85% of the entire workforce verification process.