Observe.ai Introduces 30 Billion Metric Contact Center LLM and Generative AI Product Suite

Observe.ai Introduces 30 Billion Metric Contact Center LLM and Generative AI Product Suite

Join top executives in San Francisco July 11-12 to hear how leaders are integrating and optimizing AI investments for success. Learn more

Intelligent conversation platform Observe.ai today unveiled its Large Language Model (LLM) for contact centers, with a capacity of 30 billion metrics, along with a suite of generative AI designed to improve agent performance. The company says that, in contrast to models like GPT, its proprietary LLM is trained on a vast dataset of real-world contact center interactions.

While a few similar offerings have been announced recently, Observe.ai stressed that its model’s standout value lies in the calibration and control it provides to users. The platform allows users to fine-tune and customize the model to meet specific contact center requirements.

The company said its LLM has undergone specialized training on multiple contact center datasets, equipping it to handle various AI-powered tasks (call summarization, automated QA, coaching, etc.) customized for contact center teams. contact center.

With its LLM capabilities, Observe.ai’s generative AI suite strives to improve agent performance across all customer interactions: phone calls and chats, questions, complaints, and day-to-day conversations handled by contact center teams.


Transform 2023

Join us in San Francisco July 11-12, where top executives will share how they integrated and optimized AI investments for success and avoided common pitfalls.

subscribe now

Observe.AI believes these capabilities will enable agents to deliver better customer experiences.

“Our LLM underwent extensive training on a specific dataset from the contact center interactions domain. The training process involved using a massive body of data points culled from the hundreds of millions of conversations Observe.ai has processed over the past five years,” Observe.AI CEO Swapnil Jain told VentureBeat.

Jain emphasized the importance of quality and relevance in the instruction dataset, which includes hundreds of curated instructions across various tasks that are directly applicable to contact center use cases.

This meticulous approach to curating datasets, he said, has enhanced the LLM’s ability to deliver the accurate and contextually appropriate answers the industry demands.

According to the company, its contact center LLM outperformed GPT-3.5 in initial benchmarks, showing a 35% increase in accuracy in conversation summarization and a 33% improvement in sentiment analysis. Jain said these figures should improve further through continuing education.

Additionally, the LLM has been trained on redacted data only, ensuring there is no personally identifiable information (PII). Observe.AI highlights its use of redaction techniques to prioritize customer data privacy while leveraging the capabilities of generative AI.

Eliminate hallucinations to provide accurate insights and context

According to Jain, the widespread adoption of generative AI has prompted an estimated 70% of companies across various industries to explore its potential benefits, particularly in areas such as customer experience, retention and revenue growth. Contact center leaders are among the enthusiastic users eager to take advantage of these transformative technologies.

However, despite their promises, Jain believes that general LLMs face challenges that prevent them from being effective in contact centers.

These challenges include a lack of specificity and control, an inability to distinguish between correct and incorrect responses, and limited proficiency in understanding human conversation and real-world contexts. As a result, he said these generic models, including the GPT, often produce inaccuracies and confabulations, also known as “hallucinations,” making them unsuitable for corporate environments.

“Generic models are trained on open internet data. Therefore, these models miss the nuances of human spoken conversation (think disfluencies, repetitions, broken sentences, etc.) and also struggle with transcription errors due to text-to-speech models,” Jain said. “So they could be useful for general tasks like summarizing a conversation, but lack relevant context for conversations within the contact center.”

Jain explained that his company has addressed these challenges by incorporating five years of well-processed and relevant data into its model. He gathered this data from hundreds of millions of customer interactions to train the model on specific contact center tasks.

“We have a detailed and accurate understanding of what ‘successful’ customer experiences look like in real-world contexts. Our customers can then further refine and tailor this to the unique needs of their business,” Jain said. “Our approach provides a comprehensive framework for contact centers to calibrate the machine and verify that actual results align with their expectations. This is the nature of a “glass box” AI model that offers complete transparency and builds trust in the system.”

The company’s new generative AI suite empowers agents throughout the customer interaction lifecycle, he added.

Knowledge AI feature facilitates quick and accurate responses to customer inquiries by eliminating manual searches of numerous internal knowledge bases and FAQs; while the Auto Summarize feature allows agents to focus on the customer, reducing after-call activities and ensuring the quality and consistency of call notes.

The Auto Coaching tool provides personalized, evidence-based feedback to agents immediately after concluding a customer interaction. This facilitates upskilling and aims to enhance the learning experience for agents, complementing their regular supervisor-based coaching sessions.

A new benchmark for contact center LLMs

Observe.ai says that its proprietary model surpassing the GPT in terms of consistency and relevance marks significant progress.

“Our LLM only trains on data that is completely obscured of any sensitive customer information and PII. Our redaction benchmarks for this are exemplary for the industry: We avoid excessive redaction of sensitive information in 150 million instances across 100 million calls with fewer than 500 errors reported,” Jain explained. sensitive data and respect privacy and compliance, while retaining maximum information for LLM training”.

He also said that the company has implemented a robust data protocol for archiving all customer data, including LLM-generated data, in full compliance with regulatory requirements. Each customer/account is assigned a dedicated storage partition, ensuring data encryption and unique identification for each customer/account.

Jain said we are witnessing a pivotal moment in the flourishing of generative AI. He pointed out that the contact center industry is filled with repetitive tasks and believes that generative AI will enable human talent to do their jobs with remarkable efficiency and speed, exceeding their current capabilities tenfold.

“I think successful revolutionaries in this area will focus on creating fully controllable generative AI; reliable with full visibility into results; and safe,” Jain said. “We are focusing on building reliable, trustworthy and consistent AI that ultimately helps human talents do their jobs better. We aim to create an artificial intelligence that allows humans to focus more on creativity, strategic thinking and creating positive customer experiences.”

VentureBeat’s mission it is to be a digital city square for technical decision makers to gain insights into transformative business technology and transactions. Discover our Briefings.