Director, Engineering
CurrentLeading the Search & Assist unit, part of Dialpad's Product Engineering build a best in class, search powered, virtual assistant with generative AI, semantic search and all the machine learning required.Owning a scaleable, highly mutable vector database specifically designed for powering virtual assistants with a human in the loop. Including training and hosting at scale a prompted multi-language embeddings model.Owning a cutting edge, low latency, conversational AI engine, from the early stages of R&D all the way up to product implementation and runtime operations such as live monitoring and customers onboarding.Designing and maintaining an LLM powered platform to commoditize the use of LLMs within the organization including runtime monitoring, development playgrounds, end-to-end regression testing, prompt engineering, and A/B testing of internal models and third party vendors including including OpenAI and Google Vertex.Promoting the enablement of multi language services across the organization. Including facilitating end-to-end testing across multiple languages and owning a low latency, high accuracy, conversation specific language detection service.Owning the full stack (including front end and back end) of a chatbot knowledge management platform which allows to implement the RAG flow using knowledge sourced from third party sources as well as capturing data from historical usage. Distributing a Vue3 Chatbot SDK that can be easily embedded within the Dialpad Platform to offer interactive, AI powered, chat interfaces.
