AI (Artificial Intelligence) is the phrase on everyone’s lips and minds in the software universe. And, of course, Salesforce is staying ahead of the innovation curve, releasing Einstein GPT, the World’s First Generative AI for CRM. Generative AI is the next gen development of predictive AI, where the “machine” is training itself within billions of parameters, rather than humans training AI with notable limitations. This means that AI can independently create new things from learned information.
So, what is Einstein GPT?
Einstein GPT is built around human interaction, helping the AI learn continuously as it works in tandem with service agents. The aim for this new Salesforce innovation is to reduce service costs by driving agent productivity with the help of AI. Salesforce are always leading the way for Saas enterprises, and have been investing in the benefits that AI can bring to their customers since they brought out Sales Cloud Einstein in 2016 to score leads and predict outcomes. Now, Salesforce boasts an “autonomous enterprise”, with heightened response and understanding that will elevate customer experience. Einstein will now be able to answer agent and customer questions using data from relevant sources, build landing pages, write apex code, create personalised communications and direct marketing and even live chat to customers with value adding, intuitive responses.
With a focus on ethical development and usage, Salesforce addressed four core principles when developing their new generative AI in order to empower human users with accuracy, safety, honesty and sustainability:
- Trust: human centricity in automation, verifying AI suggestions and insights for continuous improvement
- Relevance: rooted in CRM data
- Security: constructed on Hyperforce, Salesforce’s latest infrastructure, purpose built across all clouds
- Ecosystem: partnering within the space to offer freedom of choice for customers
How does it work?
Salesforce has forged partnerships with AI innovators and regulators such as OpenAI, Cohere & Anthropic to ensure the highest standards of security. Built on the LLM (Large Language Model) Gateway, Service GPT thus has a central governance solution protecting Personal Identifiable Information (PII) – therefore not exposing colleagues’ or customers’ sensitive data to open AI.
Service GPT’s mission is to fine tune CRM data with humans involved at every interaction. As the product is constructed on Hyperforce’s fully GDPR and EU Data Privacy compliant architecture, Service GPT can intuitively generate responses and answers that are even further enhanced by the insights of Data Cloud. Note that using Data Cloud is not a requirement for Service GPT. However, the more data – the more powerful your AI, and therefore your agent, will become.
Meeting customer needs with high productivity – a use case example:
Perhaps you are a Customer Service Agent working for a health and beauty retailer. A customer inquires about the products you offer for individuals with dry skin, once you accept the incoming chat bot query, Einstein Replies will make a suggestion using the available data. As the agent, you are in full control of the loop, you can adjust (rewrite reply), edit (amend manually) and accept the response to keep it appropriate and relevant to the customer exchange. A product recommendation is made, but Einstein also pulls in CRM insights using your company’s marketing data. This individual is a loyal customer who purchases your sun cream and tanning products every year at the start of Summer. Einstein suggests notifying your customer about a new, hydrating sun cream and sending them a discount code for their purchase to get them the best deal (using your Sales Cloud data).
Your customer happily ends the chat and goes to browse your new sun care range. Einstein then highlights the ‘next best action’ which may fire off a marketing campaign featuring an article on the top tips for retaining your holiday tan. You will be able to access a summary of the chat and provide feedback on the Service GPT suggestions; were they relevant, helpful and intuitive? As the agent, you even have the opportunity to create a Knowledge Article off the back of this interaction. This will allow future customers to easily identify the best SPF products for those prone to dry skin without even entering a chat! You can save the draft for the Marketing department to review and publish, further enhancing your company’s productivity.
We are so excited to see the incredible results of Service GPT when it is brought to market next month. Want to find out more about how you can personalise every customer interaction and response without sacrificing productivity? Give us a call on 0117 230 2390 or email us at firstname.lastname@example.org.