Boost.ai, a leading conversational AI solution provider, today announced Version 12 of its platform, the first of a series of planned updates by the company to incorporate Large Language Model (LLM)-enriched features. This iteration is focused on key customer experience (CX) improvements, including content suggestion, content rewriting and accelerated generation of training data. The new update will take advantage of Generative AI to suggest messaging content to AI Trainers within the boost.ai platform, generating suggested responses and resulting in drastically reduced implementation times for new intents. With this latest release, boost.ai reinforces its commitment to researching, developing, releasing, and maintaining responsible implementations of LLM-powered, enterprise-quality conversational AI features in order to further enhance the customer experience.

ChatGPT has dominated headlines throughout 2023, yet questions about the accuracy of its underlying models, GPT-3.5 and GPT-4, persist. With an 85% accuracy rate, LLMs are a good indication of the potential of this technology in consumer-facing applications, but in their current raw state they lack the dependency for integration directly into a bank or insurance firm's systems. Boost.ai helps to resolve this issue by seamlessly integrating the predictive capabilities of LLMs with enterprise-grade control of their conversational AI platform, creating a Hybrid Natural Language Understanding (NLU) system that offers unmatched accuracy, flexibility, and cost-effectiveness. Expertly weaving together the right AI components for each task, boost.ai’s platform ensures precise and accurate answers - harnessing the benefits of LLMs, while still adhering to stringent quality assurance requirements.

“LLM technology offers great promise, but most applications just aren’t properly designed to securely and scalably support real-world businesses. With worries about accuracy or even inappropriate behavior, established institutions like banks could not risk direct access to this iteration of generative AI - until now,” said Jerry Haywood, CEO of boost.ai. “By pairing LLMs with our conversational AI, we’re able to ensure accuracy and open the door for customers in sensitive industries like financial services. We’re proud to be pioneering a way forward for businesses to harness this tech right now. It’s available for customers to use and enhance their existing solution, and to help them achieve speed to value significantly sooner whilst minimizing the risks currently dominating headlines.”

With a Hybrid NLU approach, enterprises gain the ability to combine boost.ai’s market-leading intent management, context handling, and dialogue management solutions with powerful LLM-enriched tools. Boost.ai’s existing intent engine is highly trained with guardrails in place to help guide the LLM, increasing overall accuracy and reducing the number of false positives. The end result is a chatbot that can confidently provide answers to inquiries, and a more streamlined development path that radically enhances how boost.ai customers can build scalable customer experiences for chat and voice.

“With our Hybrid NLU, we’ve been able to surpass the performance of either model on its own, providing our customers with the best of both worlds,” said Lars Ropeid Selsås, Founder of boost.ai. “Boost.ai will continue to operate at the cutting edge of AI possibility, refining our platform so that our customers are always receiving the best possible technology and service.”

On May 9th, Boost.ai co-founder, Henry Vaage Iversen and the technical team conducted a webinar highlighting the strengths of Hybrid NLU and mapping out a path for businesses to start using LLM technology to their advantage. For a recording of the full webinar, please visit: https://www.boost.ai/webinars/enhancing-cx-and-reducing-operational-costs-with-boost-ai-large-language-models