OpenAI’s GPT-3 makes life easier for customer service agents

Amazing features of GPT-3, which is the largest natural language generation model available today from San Francisco-based artificial intelligence research laboratory OpenAI, had already made quite a stir on Twitter. This model uses some deep learning to generate text that is very difficult to distinguish from something written by a human, and its capabilities go well beyond text too. It can generate poems, images, programming code and even music. But I will not be re-iterating on features or technology behind GPT-3 (you can check any of these links if you’re curious). Instead, let’s focus on a single application area of this amazing tech — making customer service agent’s life easier.

If you work in customer care for a large company, or if you’re a part of the small startup team, or even solopreneur, you know how challenging daily customer communication can be. Repeated questions can take up to 80% of your time, and the worst part is that a lot of these questions require some general knowledge that goes well beyond your knowledge base or F.A.Q.

We at Activechat were lucky enough to get early access to OpenAI’s API which is in public beta now, and we attempted to put it to some good use in the customer service domain. Instead of responding to each of customers’ questions manually, your agents can now use GPT-3 to get…