Virtual Assistant Case Study
Discover how our engineers helped an edtech giant build an AI-powered virtual assistant
About the client
The client is one of the world’s largest and most diverse education providers. They offer higher education programs, professional training and certifications, test preparation and student support services. The company operates in over 30 countries, has over 10,000 employees worldwide, and thousands of customers, from universities to businesses.
About the project
The goals
Increase efficiency and productivity
The client needed to significantly speed up the work of the customer success teams, in order to handle more customer inquiries efficiently.
Reduce lead time
The goal was to build an AI-powered virtual assistant to reduce the time required for searching extensive and complex product catalogues.
The Zartis team
A dedicated team of Zartis AI Consultants started working on the project from scratch. The virtual assistant helps customer success teams by finding answers in complex documentation and enables them to deal with repetitive questions in a more efficient way.
About the project
The solution
A custom-built VA
Zartis’ AI experts delivered a custom PoC within 2 weeks. From selecting the right LMM and tech stack to designing the architectural model, our team started working on the development of a scalable virtual assistant, that can support multiple locations and languages.
Actionable data insights
The virtual assistant constantly analyses incoming inquiries and provides insights on the type, content, and the volume in order to guide future process or documentation improvements within the team.
The results
PoC in 2 weeks
Zartis AI consultants delivered a fully functional, custom Proof of Concept that ensures business requirements alignment within 2 weeks.
Easier workflows
The VA helps customer success teams by finding answers in complex documentation and answering repetitive questions more efficiently.
Multi-location support
A virtual assistant that is scalable across multiple business units, locations and languages, to support customers world-wide.
Minimised hallucinations
We restructured the data and documentation, and enabled the AI model to access and serve information more clearly.
