About our founder’s vision
I remember at some point, about 12 years ago, when many developers were talking about the fact that the rise of mobile applications would have destroyed the web. “The web is dead” event didn’t happen, as we know. The web is still here, stronger than ever (thanks, Tim Berners Lee).
Another example. The advent of chat applications didn’t destroy telephony. Now simply we have another tool, living side-by-side with the first one.
A lot of people nowadays are talking about the rise of AI, generative, and so on, and the fact that AI will steal a lot of jobs from people all around the globe. Don’t make the same mistake again.
AI is another tool as well. A great one for sure. Incredible today, normal tomorrow.
A lot of things will change, of course, and nothing will be the same as before, this is recurrent with Progress. We add opportunities. Just as an example, AI & LLMs technologies created new professions: the “Conversation Designer”, the “Prompt Designer” and the “Automation Designer”.
Here at Tiledesk, we work hard on understanding the upcoming world, before doing things (aka developing software). And AI is the best thing that could have happened for us because we’ve been on chatbots from the beginning. We believed that automating conversations was not only beneficial in the “customer service” market, but that, as sometimes happens, customer service is opening a new door on how AI can be used.
Urgent flow with ChatGPT
LLM-Enabled Conversational Apps Development: Bridging Technologies
We suddenly discovered that the Design Studio solved more problems than the ones we thought he could solve in the beginning.
Now switch our context. Move out of the customer service market and move into your company management domain (read here the article by Paolo Cervini about how ChatGPT performs in this context). Suppose you are a data analyst inside a big company and your boss asked you to get the insights he desires out of the company’s data. Today you can ask ChatGPT, using his Advanced Data Analysis features with the GPT-4 model. Submit your data as a CSV file, ask GPT to remove some rows based on some prompts, then get another CSV in reply, operate on it with another prompt to merge some columns, and so on until getting your chart in reply and finally sending it to your boss. And your boss is asking this to you…every day. It’s such a repetitive task. You need some automation to involve all these prompt chains. In the past, you should have committed a new application to your developers, designed multiple API integrations, and spent a lot of money for an application that you don’t know how will last on your Company’s toolkits.
And today? Use the Tiledesk Design Studio. Simply put some GPT blocks in a chain. Each block gets an input (a CSV) and generates an output (CSVs/Images) for the next block. Your boss will open his WhatsApp on your company’s profile, choose your chatbot (aka conversational application), and ask it for his insights. Your flow will start, connecting the blocks the way you designed them and generating your chart in your boss chat. That’s it, you just discovered a new way of building applications. You discovered a new application paradigm: the conversational app. And, unlike good old apps, it is entirely made of AI blocks, connected. Without a single line of code. Welcome to the conversational apps era.
Take a look at our article on our blog covering this vision.
Tiledesk Design Studio: A Hub for LLM-Enabled Innovation
Just think about how many applications are moving from the classic UI (web or mobile), made up of buttons and a lot of interactive, programmatically developed elements, to conversational apps.
Conversational apps are easier to develop with no code tools (like our Design Studio) because you save a lot of time skipping the UI design! The UI is…the conversation. The interaction with the UI is represented by the flowing text, built of questions and replies, along with images, videos, and buttons (that are already part of today’s chat paradigm). You only have to design the backend logic, simply connecting blocks, many of them LLMs powered.
More LLMs
The Design Studio will provide more and more LLMs engines tomorrow, not only OpenAI. Because we are already thinking of a world where there will be a lot of AI competitors (already are), each one will be special in doing something the others don’t.
More Third-party tools integrations
We’ll also add more Actions to connect with important third-party tools like important CRMs (i.e. Hubspot, Salesforce, etc.). Automating a chat has a lot of things in common with modern automation tools like Zapier and Make.com. And those tools are as powerful as the more connections they provide with the immense ecosystem of already existing and well-known SAAS applications.
More Actions
Once you have a good designer you only need more “Actions” you can do with it. We want to create a lot of Actions to allow superior features like having an integrated Database to store your collected data, the possibility to connect external low-level Databases and tools, and more importantly, the possibility to add your own Actions, to fulfill your specific Enterprise needs easily.
More community
We want to invest more in our community, which we believe is a key factor in the success of a tool like ours. This means not only powering our open source community but especially our chatbot developers community. We developed our initial version of the online Tiledesk chatbot community where everyone can share his chatbots with the world, providing ready-to-use use cases that everyone can import with just one click.
Tiledesk Community Home
And don’t forget that you can build your chatbot on the SAAS and import it in the on-premise version installed on your own data center (and vice versa). This is a Tiledesk prerogative thanks to our Open source DNA!
Conclusions
The future is shaped by innovative applications, characterized by rapid Time-to-Market (TTM), harnessing the potential of AI, and created through intuitive visual design tools initially originating from the customer service domain. These applications are then adapted into versatile horizontal tools, capable of developing a wide range of conversational applications.
Introducing Tiledesk’s upcoming positioning: evolving into a cutting-edge LLM-Enabled Conversational Apps Development Framework tailored for no-coders. Embracing an open source besides a SAAS distribution approach to minimize barriers to adopting this new development paradigm, Tiledesk will effortlessly integrate with the latest AI/LLM technologies, enabling the creation of remarkable applications for 2024 and beyond.
Urgent flow with ChatGPT
LLM-Enabled Conversational Apps Development: Bridging Technologies
We suddenly discovered that the Design Studio solved more problems than the ones we thought he could solve in the beginning.
Now switch our context. Move out of the customer service market and move into your company management domain (read here the article by Paolo Cervini about how ChatGPT performs in this context). Suppose you are a data analyst inside a big company and your boss asked you to get the insights he desires out of the company’s data. Today you can ask ChatGPT, using his Advanced Data Analysis features with the GPT-4 model. Submit your data as a CSV file, ask GPT to remove some rows based on some prompts, then get another CSV in reply, operate on it with another prompt to merge some columns, and so on until getting your chart in reply and finally sending it to your boss. And your boss is asking this to you…every day. It’s such a repetitive task. You need some automation to involve all these prompt chains. In the past, you should have committed a new application to your developers, designed multiple API integrations, and spent a lot of money for an application that you don’t know how will last on your Company’s toolkits.
And today? Use the Tiledesk Design Studio. Simply put some GPT blocks in a chain. Each block gets an input (a CSV) and generates an output (CSVs/Images) for the next block. Your boss will open his WhatsApp on your company’s profile, choose your chatbot (aka conversational application), and ask it for his insights. Your flow will start, connecting the blocks the way you designed them and generating your chart in your boss chat. That’s it, you just discovered a new way of building applications. You discovered a new application paradigm: the conversational app. And, unlike good old apps, it is entirely made of AI blocks, connected. Without a single line of code. Welcome to the conversational apps era.
Take a look at our article on our blog covering this vision.
Tiledesk Design Studio: A Hub for LLM-Enabled Innovation
Just think about how many applications are moving from the classic UI (web or mobile), made up of buttons and a lot of interactive, programmatically developed elements, to conversational apps.
Conversational apps are easier to develop with no code tools (like our Design Studio) because you save a lot of time skipping the UI design! The UI is…the conversation. The interaction with the UI is represented by the flowing text, built of questions and replies, along with images, videos, and buttons (that are already part of today’s chat paradigm). You only have to design the backend logic, simply connecting blocks, many of them LLMs powered.
More LLMs
The Design Studio will provide more and more LLMs engines tomorrow, not only OpenAI. Because we are already thinking of a world where there will be a lot of AI competitors (already are), each one will be special in doing something the others don’t.
More Third-party tools integrations
We’ll also add more Actions to connect with important third-party tools like important CRMs (i.e. Hubspot, Salesforce, etc.). Automating a chat has a lot of things in common with modern automation tools like Zapier and Make.com. And those tools are as powerful as the more connections they provide with the immense ecosystem of already existing and well-known SAAS applications.
More Actions
Once you have a good designer you only need more “Actions” you can do with it. We want to create a lot of Actions to allow superior features like having an integrated Database to store your collected data, the possibility to connect external low-level Databases and tools, and more importantly, the possibility to add your own Actions, to fulfill your specific Enterprise needs easily.
More community
We want to invest more in our community, which we believe is a key factor in the success of a tool like ours. This means not only powering our open source community but especially our chatbot developers community. We developed our initial version of the online Tiledesk chatbot community where everyone can share his chatbots with the world, providing ready-to-use use cases that everyone can import with just one click.
Tiledesk Community Home
And don’t forget that you can build your chatbot on the SAAS and import it in the on-premise version installed on your own data center (and vice versa). This is a Tiledesk prerogative thanks to our Open source DNA!
Conclusions
The future is shaped by innovative applications, characterized by rapid Time-to-Market (TTM), harnessing the potential of AI, and created through intuitive visual design tools initially originating from the customer service domain. These applications are then adapted into versatile horizontal tools, capable of developing a wide range of conversational applications.
Introducing Tiledesk’s upcoming positioning: evolving into a cutting-edge LLM-Enabled Conversational Apps Development Framework tailored for no-coders. Embracing an open source besides a SAAS distribution approach to minimize barriers to adopting this new development paradigm, Tiledesk will effortlessly integrate with the latest AI/LLM technologies, enabling the creation of remarkable applications for 2024 and beyond.
[/vc_column_text]
It’s not uncommon that someone working in a field ends up finding a solution for another, unexpected, totally different field. Just read this list of NASA spinoff technologies (did you know about the story of CMOS technology?)
[/vc_column_text][/vc_column][/vc_row]
Empowering Customer Service with LLM-Enabled Conversational Apps
With the advent of mainstream chat applications, instant messaging emerged as a must-have tool to communicate with your customers. Today’s Customer service instant messaging is made up of People and Chatbots working together. And this was the mantra of Tiledesk since day one.
Let’s look at a typical use case, the end-user asking for support. There is often a chatbot in front of the communication. The user asks and the chatbot replies using the AI/LLMs-powered Knowledge Bases. The chatbot can then continue to “flow”, for example when he is unhappy with the automated reply and wants to connect to a human. Before connecting the end-user and the operator, the chatbot will ask for some data, show the privacy policy for acceptance, etc. Sometimes user data is also sent to an external CRM like Hubspot, and sometimes the chatbot gets back to the user with data obtained through some Company’s APIs.
How you build today is much different from how you were building yesterday. We switched from an intent-based design approach to a block-based design. A super-intense use of a programmatic approach in complex intent replies moved to a visual, no-code approach for replies development. And this happened in the end, chatbots met the no-code movement using a lot of the visual graph representation of complex logic, based on concepts like nodes, blocks, actions, conditional branching, etc.
Tiledesk Design Studio
Automate by connecting nodes
Chatbots moved from intent-based to block-based. A block is a piece of logic that can be interconnected in various ways to other blocks, rendering your chatbot complex logic.
For the no-code or citizen developer, it is very important to get a flexible tool that uses the same visual language that he uses to draw logic on a piece of paper
Sketch of a chatbot flow
Intents were not enough for these kinds of complex flows. Intents were easy to adopt when creating a question-answer bot but suddenly found their limits when chatbots increased in complexity.
This is the main reason why we developed our visual designer.
The designer is built around the concept of a super-node, meaning a visual block on a graphical stage that can do many tasks (Actions) and connect to other blocks in a non-sequential flow to implement the chatbot logic design.
Now the conversation developer no longer needs a piece of paper to sketch the flow and submit it to the Development department. Just sketch & deploy your chatbot by yourself, with only one tool with a drag-and-drop UX. Yes, the UX, it’s the most important element. The developer must be happy to solve problems through an engaging environment, not only flexible, intuitive, and performant, but also beautiful and exciting!
As a side note, the designer was created from scratch, without third-party tools or libraries, to make it available to everyone (yes, Tiledesk is open source, MIT license, and so is our Visual Designer on Github – leave us a ⭐️).
Moving to the Future: LLM-Enabled Conversational App Development
Now let’s think about the future, inspired by today’s experiences.
We discovered that providing logic to a block gives the conversation developer more power in creating their flows. What happened when we integrated chatGPT into the flow? Immediately, we discovered an incredible boost for the no-code development. Initially, ChatGPT was embedded to create fast, generative replies for the end-users. But suddenly our customers started to use ChatGPT to analyze a user phrase and make decisions based on his replies. With a chatGPT block, you can simply ask “Tell me if this phrase is about an urgent task”. ChatGPT replies. And you can branch your flow accordingly. Et voilà, we moved from the initial supernode to the AI node, able to analyze a sentence and move to the next block accordingly.
Urgent flow with ChatGPT
LLM-Enabled Conversational Apps Development: Bridging Technologies
We suddenly discovered that the Design Studio solved more problems than the ones we thought he could solve in the beginning.
Now switch our context. Move out of the customer service market and move into your company management domain (read here the article by Paolo Cervini about how ChatGPT performs in this context). Suppose you are a data analyst inside a big company and your boss asked you to get the insights he desires out of the company’s data. Today you can ask ChatGPT, using his Advanced Data Analysis features with the GPT-4 model. Submit your data as a CSV file, ask GPT to remove some rows based on some prompts, then get another CSV in reply, operate on it with another prompt to merge some columns, and so on until getting your chart in reply and finally sending it to your boss. And your boss is asking this to you…every day. It’s such a repetitive task. You need some automation to involve all these prompt chains. In the past, you should have committed a new application to your developers, designed multiple API integrations, and spent a lot of money for an application that you don’t know how will last on your Company’s toolkits.
And today? Use the Tiledesk Design Studio. Simply put some GPT blocks in a chain. Each block gets an input (a CSV) and generates an output (CSVs/Images) for the next block. Your boss will open his WhatsApp on your company’s profile, choose your chatbot (aka conversational application), and ask it for his insights. Your flow will start, connecting the blocks the way you designed them and generating your chart in your boss chat. That’s it, you just discovered a new way of building applications. You discovered a new application paradigm: the conversational app. And, unlike good old apps, it is entirely made of AI blocks, connected. Without a single line of code. Welcome to the conversational apps era.
Take a look at our article on our blog covering this vision.
Tiledesk Design Studio: A Hub for LLM-Enabled Innovation
Just think about how many applications are moving from the classic UI (web or mobile), made up of buttons and a lot of interactive, programmatically developed elements, to conversational apps.
Conversational apps are easier to develop with no code tools (like our Design Studio) because you save a lot of time skipping the UI design! The UI is…the conversation. The interaction with the UI is represented by the flowing text, built of questions and replies, along with images, videos, and buttons (that are already part of today’s chat paradigm). You only have to design the backend logic, simply connecting blocks, many of them LLMs powered.
More LLMs
The Design Studio will provide more and more LLMs engines tomorrow, not only OpenAI. Because we are already thinking of a world where there will be a lot of AI competitors (already are), each one will be special in doing something the others don’t.
More Third-party tools integrations
We’ll also add more Actions to connect with important third-party tools like important CRMs (i.e. Hubspot, Salesforce, etc.). Automating a chat has a lot of things in common with modern automation tools like Zapier and Make.com. And those tools are as powerful as the more connections they provide with the immense ecosystem of already existing and well-known SAAS applications.
More Actions
Once you have a good designer you only need more “Actions” you can do with it. We want to create a lot of Actions to allow superior features like having an integrated Database to store your collected data, the possibility to connect external low-level Databases and tools, and more importantly, the possibility to add your own Actions, to fulfill your specific Enterprise needs easily.
More community
We want to invest more in our community, which we believe is a key factor in the success of a tool like ours. This means not only powering our open source community but especially our chatbot developers community. We developed our initial version of the online Tiledesk chatbot community where everyone can share his chatbots with the world, providing ready-to-use use cases that everyone can import with just one click.
Tiledesk Community Home
And don’t forget that you can build your chatbot on the SAAS and import it in the on-premise version installed on your own data center (and vice versa). This is a Tiledesk prerogative thanks to our Open source DNA!
Conclusions
The future is shaped by innovative applications, characterized by rapid Time-to-Market (TTM), harnessing the potential of AI, and created through intuitive visual design tools initially originating from the customer service domain. These applications are then adapted into versatile horizontal tools, capable of developing a wide range of conversational applications.
Introducing Tiledesk’s upcoming positioning: evolving into a cutting-edge LLM-Enabled Conversational Apps Development Framework tailored for no-coders. Embracing an open source besides a SAAS distribution approach to minimize barriers to adopting this new development paradigm, Tiledesk will effortlessly integrate with the latest AI/LLM technologies, enabling the creation of remarkable applications for 2024 and beyond.
