More conversations with less effort: Extending possibilities with GPT-powered Chat

AI has been on Efecte’s radar for years. Our AI journey started with the release of a Virtual Coach, an AI-based digital assistant that helps reduce agents’ workload. We then acquired InteliWISE, a Polish-based company with strong expertise in the Conversational AI space, to bring our solution capabilities to a new level. And it does not end there.

Today, we are all witnessing a generational leap in AI capabilities. Initiated by OpenAI's launch of the ChatGPT product, we have seen an unprecedented take-up of the new AI language models and applications leveraging them. At Efecte we are constantly finding ways to leverage the new technological advances for the benefit of our users, which led us to the most recent decision of embarking on the GPT journey. With this, we are proud to announce the release of the GPT-powered Chat. This is the first step towards our Effie AI family, which will leverage AI technology to boost user efficiency across all Efecte solutions while keeping human in control.

What does GPT mean for Service Management?

Generative Pre-trained Transformer (GPT) is a language model that is trained to create human-like conversations. Generative models are largely used by businesses today in the form of chatbots, virtual assistants, and content generators. They communicate with users in natural language, providing personalized recommendations and assistance, and create engaging and informative content quickly and efficiently.

When it comes to service management, GPT-3 and GPT-4 language models have the potential to enable more effective communication between ITSM professionals and end-users, allowing for faster problem resolution and improved customer satisfaction. With their processing and analyzing capabilities, GPT models could help ITSM teams identify patterns and trends, allowing for more proactive problem-solving and service improvements. Overall, they play a big role in improving the efficiency and effectiveness of ITSM operations.

How can GPT affect the key metrics in IT Service Management?

One of the biggest challenges in ITSM and Service Management nowadays is the universal need to reduce costs per agent-handled contact. Operationally, Service Desk teams are looking for ways to improve two important Service Desk KPIs: Response Time (RT), and Average Handle Time (AHT). This is where GPT model comes into play. While relying on the support of GPT to answer users' tickets, agents should be able to cut the response time significantly. This, in turn, will affect other important metrics, such as incident resolution times, service request fulfilment, and user satisfaction.

GPT-powered Chat for improved quality and timeliness of responses

With GPT-powered Chat, Service Desk agents will be able to compose responses to users’ inquiries based on automatically generated prompts in a split second, while always remaining in control. With its superpowers, Service Desk agents can:

Autogenerate

When an agent opts in to use GPT-powered Chat, the technology generates a response that is relevant to the context of user inquiry. The GPT-powered functionalities can be combined with other data from the customer's Efecte platform, such as the workstation model, which will allow even more targeted answers to the end user. Agents can then check and verify the response and decide on the final answer being sent. With GPT-powered Chat support, this process is fast and efficient, allowing agents to provide high-quality service in a timely manner in any language.

Autocorrect

Agents can quickly type out a reply, and then replace casual language and grammatical errors with a more formal or correct alternatives. This feature allows agents to focus on capturing the essence of the message while GPT-powered Chat ensures the clarity and cohesiveness of communications. It is especially beneficial for agents who are new or are not proficient in the local language.

Autocomplete

GPT-powered Chat predicts sentence completion, helping agents save keystrokes and deliver a faster response to end-user questions​. Avoiding errors leads to more accurate and standardized communication, which impacts the customer support quality.

Addressing ethical and legal concerns

Right now, the use of generative models is surrounded by a number of concerns, including misinformation and data leakage. With our customers’ and users’ safety being at the heart of our business, we would like to proactively address these worries.

All data provided to GPT will be sent through their API. According to OpenAI Terms of Service, data sent through the API will not be used for training their algorithms or otherwise improving their services. This means customer data cannot leak to other GPT users. The use of GPT-powered Chat is fully legal in all Efecte geographies. Commercial use is allowed under applicable laws and OpenAI terms. We’re following legislation and authorities’ decisions and will make adjustments if needed.​

We are keeping human at the center of this extension, giving Service Desk agents the possibility to choose when to opt-in for the use of GPT-powered Chat and the responsibility to decide on the final message that is being sent. This allows us to tackle any ethical issues and guarantee that all communications follow ethical and business standards.

Interested to discuss how to empower ITSM with GPT?

We are excited to announce that GPT-powered Chat is now ready for piloting! We've been working hard to develop this new feature, and we're eager to share it with our customers. We invite all of our customers to test it out and provide feedback.

Your input is very valuable to us, and it will help us make any necessary improvements before we launch the feature to a wider audience. We appreciate your participation and look forward to hearing your thoughts.

Are you willing to give it a try? Talk to us!