Unlike generic models, Observe.AI’s LLM is trained on real-world contact center data and tasks, and offers greater accuracy, control, and privacy for contact centers to leverage Generative AI
Observe.AI, the live conversation intelligence platform for contact centers, today introduced its 30-billion-parameter Contact Center LLM and new Generative AI Suite for boosting agent performance.
Unlike generic models like GPT, Observe.AI’s proprietary large language model (LLM) is trained on a domain-specific dataset of hundreds of millions of customer interactions. This enables it to support a diverse set of AI-based tasks that are highly specific to contact center teams. Observe.AI’s new Generative AI Suite leverages this LLM to enhance agent performance before, during, and after customer interactions – through capabilities like surfacing real-time answers, automated call summarization, and automated coaching notes that drive agent self-improvement.
“We’re at an exciting precipice for the use of generative AI in contact centers – an inflection point on par with the advent of the cloud or mobile. It’s a critical moment that will separate the disruptors from the disrupted, and contact centers who move forward with LLM strategies based on accuracy, calibration, and control will realize their fullest potential,” said Swapnil Jain, CEO and Co-Founder of Observe.AI.
“By leveraging a domain-specific LLM, we’re able to drive deeper trend analysis, more accurate call summarization, and in-context question answering while ensuring degrees of control, calibration, and privacy that are simply not possible with generic models”
Why Do Contact Centers Need a Domain-Specific LLM?
With the generative AI boom afoot, it’s estimated that 70% of businesses across industries are currently in exploration mode – citing customer experience, retention, and revenue growth as top focus areas. Contact center leaders are among those eager to jump in and harness these technologies.
But while generative AI holds massive promise, there are several challenges to using generic LLMs that dampen their effectiveness in contact centers. They include a fundamental lack of specificity and control, inability to discern right from wrong responses, and ineptitude with spoken human conversation and real-world environments. Consequently, generic models like GPT are prone to serious inaccuracies and confabulations – otherwise known as “hallucinations” in AI – making them too risky to use in business settings.
Greater Accuracy, Control, and Privacy with Contact Center LLM
Observe.AI’s Contact Center LLM delivers higher accuracy and control through 5+ years of human calibration and feedback. As the model is already fine-tuned to the unique needs of the contact center, it offers higher performance out-of-the-box compared with generic models. It can also be further refined to target the customer’s specific business objectives, needs, or use cases.
“By leveraging a domain-specific LLM, we’re able to drive deeper trend analysis, more accurate call summarization, and in-context question answering while ensuring degrees of control, calibration, and privacy that are simply not possible with generic models,” said Vache Moroyan, SVP of Product at Observe.AI.
Initial benchmarks demonstrate that Contact Center LLM is 35% more accurate than GPT3.5 in automatically summarizing conversations and 33% more accurate in sentiment analysis. These numbers are expected to improve with continuous training. Additionally, the LLM is trained only on data that is completely redacted of any Personally Identifiable Information (PII) – using the industry’s most accurate redaction techniques – ensuring customer data privacy while using generative AI.
Accuracy and flexibility are major differentiators that draw companies like Public Storage, Bill.com, and Cox Automotive to Observe.AI’s platform. Observe.AI customer Accolade, a leader in the healthcare industry, prioritizes a high-touch, personalized experience when it comes to member engagement. Accolade’s frontline healthcare teams are highly specialized and require a level of calibration that elevates their standard of care, consistently and at scale.
“It’s critical for us to partner with technology solutions that are innovative, forward-thinking, and lean into emerging technology,” said Ardie Sameti, Senior Director of AI & Automation at Accolade, Inc. “Observe.AI’s LLM empowers us to proactively engage our members with precise data about their care needs while managing delivery of empathetic and personalized experiences. Generative AI solutions like Auto Summary help us save significant time in after-call work and allow our front line healthcare team to focus on building lasting relationships with every member we serve.”
Read More: Miva, Inc. Announces The Release Of New Miva Version 10.07
Observe.AI’s New Generative AI Suite for Agent Performance
New generative AI capabilities announced today help contact centers boost agent performance before, during, and after customer interactions, including:
- Knowledge AI: Saves the time agents spend manually searching knowledge bases and FAQs by providing answers to customer questions – increasing first call resolutions and reducing AHT
- Auto Summary: Automatically and consistently captures interaction summaries in multiple formats – structured, unstructured, and entities – eliminating after-call work, allowing the agent to focus on the customer, and improving the quality and consistency of notes
- Auto Coaching: Automatically generates and serves up coaching notes for agents on the spot as soon as a customer interaction ends – driving immediate skills improvement, better CX, and faster feedback for performance improvement in addition to regular supervisor-assisted coaching