Dialpad Launches Large Language Model for Enterprises of All Sizes

Dialpad has unveiled DialpadGPT, a domain-specific large language model (LLM) to power features that automate tasks for enterprises regardless of size. Dialpad, an AI-powered customer intelligence platform, built the model using more than five years and five billion minutes of proprietary conversational data.

This real-time generative AI is based on the language of business and utilizes conversational and messaging data across customer service, sales, recruiting and employee collaboration. It harnesses the power of AI automation and AI assistance at an unprecedented scale, Dialpad officials said.

While existing LLM-based models like ChatGPT have proven to be a gateway AI for businesses, substantial challenges to adoption persist including limitations of scale and prevailing concerns about accuracy, lack of business-specific data, privacy and security. DialpadGPT addresses these issues.

Dialpad’s Dan O’Connell

Dan O’Connell, chief AI and strategy officer at the company, said LLMs are only as accurate and as strong as the data they are trained on.

“DialpadGPT was purpose-built for the enterprise, and addresses businesses’ common qualms with existing LLMs. It delivers unparalleled accuracy, produces unbiased results at a fast speed, provides the opportunity to scale, and meets high security and compliance standards — all at a lower cost than existing LLMs. AI has been at the heart of Dialpad’s business for more than five years, and we’re excited to bring such a transformative tool to market.”

Large Language Model Making AI Readily Available

Despite the widespread acknowledgment of generative AI’s potential to reshape the business landscape, its adoption and implementation pose considerable challenges due to limited global capacity, including a shortage of computational power and escalating costs that make AI transformation an uphill task. DialpadGPT, designed with scalability in mind, makes AI readily available and enables enterprises to take advantage of Dialpad’s generative features. For businesses grappling with AI and machine learning (ML) application deployment due to knowledge gaps or a lack of internal expertise, DialpadGPT presents no-code options that effectively address the hurdle to AI uptake in sales, service and recruiting. 

DialpadGPT is a further evolution of Dialpad’s AI technology stack. The stack features a proprietary real-time speech recognition (ASR) engine, natural language processing (NLP) models and semantic search capabilities. It is capable of boosting worker productivity and driving a deeper understanding of customer attitude and sentiment and transforming how teams work with their customers.