Voice Tech Podcast
Published in

Voice Tech Podcast

Design with me: Chatbot Design

Alfred, a VUI-based Bullet Journaling Chatbot for Google Assistant Platform

In this article I am going to capture my process on creating a chatbot named ‘Alfred’. He is a bullet journaling chatbot.

He will help user to create bullet journal in their day to day life. Let’s design, learn, relearn, fail and prevail together. 🙂

I have recently started doing bullet journaling and I have found out it is very helpful for me and it makes me very productive when I have my goals/tasks for the day already written out.

Step 1: Goal for Alfred

Alfred will help me add tasks in my bullet journal. Adding, deleting, delegating, moving tasks and events manually everyday takes some quality time. I have found that I often forget to plan and add tasks in my Bullet Journal when I spend a hectic day at work. I have also dedicated a time (after brushing my teeth at night) for it. However, I feel too lazy to write after a long day.

I simply want an assistant who will listen and record and maintain the tasks in the journal on behalf of me 👸.

Step 2: User and Chatbot Persona

User Persona

Vijaya, 29, an Interaction designer in Walmart. She is working with multiple domains and runs couple of initiatives in her day to day work life. She also loves to cook. Hence, she and her husband take turn and cook by themselves.

It gets pretty hectic for her to juggle work, cook healthy homemade food and also work on the creative ideas/initiatives. She has started bullet journaling to keep track of everything and set her work-life priorities.

Chatbot Persona

Alfred, 33, a male. He is helpful, easy to approach and a motivating person. He is not too playful, not too serious.

He is attentive and focused. Someone whom you can count on. You can go into more details while designing the personas, like how we do for GUI apps with motivations, behaviors, likes, dislikes etc.

I am keeping this as simple sentences for now.😇

Step 3: Use Case

Okay now we have our personas ready, let’s focus on some key use cases. Keep them simple and to-the point for now. We are learning how to do conversation design together. Short and simple. So here we go, below are few key use cases for using Alfred.

1. Add task for next day

2. Add a task for current day

3. Mark a task as complete

4. Cancel a task

5. Move a task to next day

I have listed 5 use cases, however based on the goal of the chatbot you are choosing you may find more than 5. For this exercise let’s choose the no. 1 use case which is ‘Adding task for next day’.

Step 4: Sample Dialogs

Now the fun stuff. Get pen and paper ready or just go to the whiteboard and start writing few sample dialogs. Let me show you how I have written few sample dialogs between Me(user) and Alfred(bot) for ‘Adding task for next day’ use case.

Okay, now let’s analyze what I have done in each flow.

Flow 1: Adding Task for Tomorrow

Remember: The Cooperative Principle by Paul Grice. Read my article on Conversation Design to know more on this.

Flow 2: Error/Exception Scenario

Build better voice apps. Get more articles & interviews from voice technology experts at voicetechpodcast.com

Flow 3: Additional User Input

Step 5: Flow Map

We are ready with personality, tone and sample dialogs for our use case ‘Adding task for next day’. Let’s dive little more into details with a flow diagram. It will help us to have an overall view of how the conversation should flow to help user achieve her goal and along the way what are the probable scenarios we can face and how shall we handle to redirect the user back to the conversation.

Before starting let us get familiarized with some symbols (legends):

Flow Map for Adding task tomorrow & Fallback

Step 6: DialogFlow

There are couple of application where we can design our conversation and test it out. I have used Dialogflow for Google Assistant platform. You can use https://www.voiceflow.com/ or https://botsociety.io/ for the design.

Intents, Training Phrases and Bot Response
Okay so to start off on the dialog flow we need to know about Intent, Training Phrases and Bot Response. Here the intents we have captured are-

1. Day Selection

2. Add Task

3. Welcome

4. Outro

5. Fallback

For Add Task Intent let’s see what are the Training Phrases we can add-

Now let’s add what the Bot should respond based on the Training Phrases. Key is to have multiple response. I cannot stress enough on this. It will help our bot to behave more humanly rather than a broken record🙅‍♀‍

You can add context and events to the intents as well. There are many good tutorials on Dialogflow on Youtube. You can go through them to learn about Dialogflow console.

Final Prototype

Hope you will find the above steps handy while starting your journey for Conversation Design and create voice based or text based chatbots. I would love to see if you guys have created anything on conversation design. If you want to connect with me on this or just to say hi, feel free to reach out to me at https://www.linkedin.com/in/thegr8madgirl/ 🙂

Thank you for reading this article. If you are happy and you know it, you can give some clap 😇.

Something just for you

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Vijaya Das

Interaction Designer @Google, a traveller at heart, finds cooking theraputic😌, obsessed about Minions and Ghibli movies✨, ice breaker 🙌