• Fri. Jun 14th, 2024

Simplr unveils Cognitive Paths to enable safe use of ChatGPT for customer service


May 11, 2023
Simplr unveils Cognitive Paths to enable safe use of ChatGPT for customer service


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

Customer experience (CX) solutions platform Simplr announced today the launch of Cognitive Paths, a new generative AI technology with comprehensive safeguards to enhance customer service interactions. By incorporating OpenAI’s ChatGPT into its platform, Simplr’s Cognitive Paths empowers customer service organizations to fully harness the power of large language models (LLMs) and generative AI while eliminating associated risks to customer experience and brand reputation.

The solution reduces the overall volume of information available to the LLM-powered chatbot and eliminates the chance of hallucination, i.e., engaging in off-brand topics. Furthermore, enterprise-grade security protocols ensure that no data is transferred back into any publicly available LLM, thereby safeguarding sensitive customer data.

“Simplr’s Cognitive Paths works by generating unique and proprietary data sets for each client,” Eng Tan, CEO and founder of Simplr, told VentureBeat. “The result is an extensive curated knowledge base, but a vastly more targeted and customized set of data than what ChatGPT has access to on its own. Cognitive Paths then directs the LLM to only pull information from certain data sets based on the nature of each unique customer interaction. The result is the effectiveness of generative AI without the hallucinations.”

Simplr has curated an extensive knowledge base consisting of data sets such as knowledge base content, product collateral, top-rated human resolutions, brand policies, and their own experience in the customer support and service space. This knowledge base is more targeted and customized compared to the data that ChatGPT has access to on its own.


Transform 2023

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


Register Now

“Using the combined power of LLMs with Simplr’s extensive curated data set of what makes excellent customer service interactions in the industry, we can now essentially automate that first-rate experience for our clients across a wide array of customer service inquiries,” said Simplr’s Tan. “The result is better brand reputation, more return customers, and more revenue generated through customer service than ever before.”

The company considers the launch of Cognitive Paths a significant milestone in the customer service industry. It believes that this technology will enable businesses to automate complex inquiries, provide superior customer experiences, and gain a competitive edge over their peers.

“Other chatbot providers have historically focused on easy, transactional questions that usually take one or two steps to complete,” Tan said. “With Cognitive Paths, we have injected our experience of more complex human interactions, especially in technical support. It means we can now automate, via LLMs, those types of questions that are still outside the purview of other chatbot providers.”

Eliminating hallucinations to provide true insights

According to a study by Juniper Research, AI will automate 70% of customer inquiries by the end of 2023. However, using generative AI in customer interactions also carries risks that can harm the brand reputation and customer relationships if not properly managed. Generative AI can often generate irrelevant content and engage in off-brand topics, making its use inherently risky.

“The primary philosophy behind Cognitive Paths is database segregation for safety. Generative AI technologies hallucinate because of the amount of data they have access to. Out of the box, Large Language Models (LLMs) cannot autonomously discern the accuracy and authenticity levels of knowledge bases, which can result in incorrect or nonsensical answers,” Simplr’s Tan told VentureBeat.

Simplr’s new set of safeguards includes OpenAI’s ChatGPT and a set of AI-training parameters to guide the chatbot technology toward the correct customer resolution without the risk of hallucination. This addresses the potential risk of negative impacts on brand reputation and customer relationships arising from generative AI in customer interactions.

Cognitive Paths utilizes enterprise-grade security protocols to ensure no back data transfer into any publicly available LLM, thereby eliminating any potential risk of disclosing PII or sensitive customer data. Additionally, by leveraging data accumulated from Simplr’s history of optimizing its outsourced human workforce, Cognitive Paths significantly broadens the range of customer interactions that can be effectively resolved through automation.

“Using generative AI in customer-facing situations without the proper guardrails is a recipe for a PR or security disaster. At the same time, it can open up data security liability,” explained Tan. “On the customer service front, using uninhibited LLMs means you also open up a very real risk that sensitive customer data can also be exposed. So, in addition to data restriction, we have implemented enterprise-grade security protocols in Cognitive Paths, which prevent PII or sensitive customer data from being shared back into the LLM data pools.”

Tan stated that his company has been dedicated to understanding the range of multi-turn inquiries humans can respond to since the beginning. This encompasses a broader and varied set of inquiries than conventional chatbot technology providers typically address.

In addition, it covers complex technical support scenarios as well as upsell and cross-sell actions.

“Other chatbot providers have always thought “What can be deflected?” Meanwhile, Simplr has always focused on “What can be resolved?” “What replicates the very best human agents?” That is what gives us a distinct advantage in the customer service generative AI race,” he said.

Leveraging OpenAI models to enhance customer interactions

The platform leverages GPT 4.0’s generative capabilities to improve customer interactions. 

According to Tan, LLMs can now summarize vast amounts of unstructured data, which was impossible with previous NLP technologies. Frequently, customers include extraneous or irrelevant information in their inquiries, but LLMs can sort through this noise to identify the actual issue and succinctly summarize it for customer confirmation before proceeding. This feature also has significant applications in agent-assist tools.

“We are particularly excited about GPT 4’s multimodal capabilities, such as image recognition. This is a big step forward in terms of engaging customers via chat. But, again, think of technical troubleshooting,” he added. “Now, in determining the make and model of a particular device, a customer can simply take a picture of it and send it to the bot. Previously, the customer would have to take time to find that information either on the device itself, in the user manual, in online product information, etc.”

He argues that although LLMs can add personality to other chatbot offerings or help locate the correct answer more quickly, they do not fundamentally expand the range of tickets that can be automated. In contrast, with Cognitive Paths, users could now automate much more complex and multi-turn inquiries.

“Think warranty and insurance claims, technical support, or automating product inquiries for huge retailers and OEMs with a plethora of products. One of the things I’m most excited about is the ability to automate conversational commerce,” said Tan. “This will make it much easier and cost-efficient for customer service teams to start driving revenue than in the past.”

Regarding the safety and security of customer data, Tan emphasized that Simplr has always handled customer and consumer data with utmost care. The company has implemented enterprise-grade security protocols to ensure the safety of sensitive customer data and personally identifiable information (PII).

“We are taking advantage of a host of tools and protocols from Microsoft Azure, dedicated environments, and our parent company, Asurion, to maintain the highest possible standards for enterprise security,” explained Tan. 

What’s next for Simplr? 

Tan believes that AI is revolutionizing customer service, and no aspect of the business will be more influenced by AI than customer service and experience. The archaic call center and BPO models from the past are not suitable for a digital and automated future.

“Everyone has had that experience with a first-rate customer service agent. They are friendly, talk to you like a human, and can resolve your issue quickly. But, unfortunately, the vast majority of customer service inquiries are forgettable at best and most often frustrating and slow,” he said. “When we started Simplr, we set out with a straightforward mission: “How can we scale first-rate customer service interactions?” And we built a platform that turned each of our human agents into the best possible agent for our clients at any given time.”

He explained that Simplr envisions the future of customer service as a seamless integration of all aspects of the customer experience related to support interactions. The company plans to leverage the incredible power of generative AI to deliver equally exceptional experiences whether customers are interacting with a bot or a human and even during in-person support interactions.

“The platform we’ve built enables this integrated experience for customers across the support spectrum,” Tan added. “Despite how far we’ve come since the launch of ChatGPT last fall, this is only the beginning. There is so much white space about where we can collectively take the customer experience over the remainder of the decade. Simplr will play a leading part in facilitating that change.”

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.

Leave a Reply

Your email address will not be published. Required fields are marked *