ChatGPT as a Workflow node

In October Tiledesk released his new Visual Flow Designer (aka Chatbot Design Studio).

The idea to have a visual designer is not new. The visual designer paradigm is shared by a lot of domains. It’s a common tool when a professional wants to reach a goal that he could anyway achieve with coding but he doesn’t want to code.

 

These professionals are part of the modern concept of “Fusion Teams” (Gartner article) often available as a borderless workforce in the modern – post covid – Enterprise.

 

As of this Gartner article we can say that “Organizations are increasingly turning to low-code development technologies to fulfill growing demands for speed application delivery and highly customized automation workflows” 

In all modern Visual Flow Designers you manage concepts like nodes, actions and branches. They are all the same. The visual tools only differ for the domain where all this stuff is applied. We, in Tiledesk, are in the conversational market, and we developed our flow designer to enable users to easily manage conversational automations (aka chatbots) development.

Leveraging ChatGPT as Your Conversational Engine

Today everyone is aware of the chatGPT power. Prompt design is a de-facto new skill. And his power is all to be discovered, because all the generative AI technology is improving day by day.

 

How do prompt design and generative AI matter with visual designers? A lot. With a no-code conversation editor you can put your prompt at work to reach more goals than a simple “super intelligent generated content” as reply.

 

An example. Put a node in the flow, enabled with ChatGPT. Ask ChatGPT to reply, taking advantage of his unique ability to evaluate a user provided sentence. In particular, ask to evaluate a sentence based on his perception of “urgency”, generally similar to the one of a human being (NDR: your perception). But – and here is the point – ask chatGPT to generate a reply not for humans but for the machine. In this way we can apply a conditional block immediately after, and continue the flow according to the perception of urgency provided by ChatGPT, based on our prompt.

gpt bot design studio

The main node is the fallback block where the ChatGPT action evaluates a prompt with the user text:

chatgpt prompt for the chatbot

Write the right prompt and ask GPT to only get “low” or “high” as a reply. You can evaluate the reply in the next condition block:

condition for chatgpt bot

Based on “urgency” evaluation (low or high) we’ll follow a high priority or a low priority path.

Final Thoughts: ChatGPT’s Role in Automation

When you use ChatGPT in a flow you can think of it like an intelligent “node” or, better, an intelligent gear in the (automation) engine. It switches the prompt design from the “generate for the humans” to “generate for the machine” perspective.

 

Where the flow is the machine, ChatGPT is his “adaptive gear”. It analyzes the conversation attributes (like user text) and “branches” the flow, it makes decisions and moves the automation-machine in the right direction. It becomes the best horizontal “no-code node” you can get today in a flow designer.

In conclusion, Tiledesk’s recent introduction of the Visual Flow Designer marks a significant milestone for developers and conversation designers in the realm of chatbot and conversational automation development. This tool speaks to the needs of the modern collaborative workspace, often featuring diverse professionals, where coding isn’t always the preferred route.

 

As organizations increasingly embrace low-code development technologies to meet the demand for rapid application delivery and highly tailored automation workflows, this Visual Flow Designer is a game-changer. It puts the power of chatGPT and prompt design at your fingertips, revolutionizing the way you approach conversational automation.

 

The synergy between visual flow design and chatGPT’s generative AI technology is particularly intriguing for those in conversation and prompt design. With this tool you can instruct chatGPT to craft responses not for humans but for machines, opening up an entirely new realm of possibilities for creating automation workflows.

 

For conversation designers, this means that you can create highly customized, context-aware interactions that provide a more personalized and efficient user experience. Developers can leverage this technology to streamline the development process, making it easier to implement complex conversational logic without extensive coding.

 

In essence, Tiledesk’s Visual Flow Designer empowers developers, conversation designers, and prompt designers to harness the full potential of chatGPT’s generative AI capabilities. It’s a powerful tool that facilitates more efficient and tailored conversations and interactions, ultimately enhancing the user experience and advancing the field of conversational automation.

Ready to revolutionize your conversational automation?

Sign up for a free trial of Tiledesk’s Design Studio today!

Andrea
Andrea
Tiledesk founder. Coding chatbots as hobby. Working with my team to create the best Conversational Apps Development Platform - aka Tiledesk

Leave a Reply