Scribd
Upload
28 views2 pages

Task - ChatBot

Uploaded by

Aditya Chourasiya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
28 views2 pages

Task - ChatBot

Uploaded by

Aditya Chourasiya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

VoiceOwl

Bots

Assignment: Advanced conversational chatbot for data gathering and deployment

Objective:
Inspired by Typeform's chat feature, especially Formless AI's user data collection approach, we
aim to create an interactive conversational chatbot that actively engages users while discreetly
handling data collection.

This chatbot should be a persuasive conversationalist, encouraging users to share their


information willingly. For example:

User: "I'm unsure if I want to share my email."


Chatbot: "I completely understand your concern. Your email is safe with us and will only be
used for sending updates on our services and exclusive offers. Is there anything specific you're
concerned about?"
User: "I'm not comfortable sharing my phone number."
Chatbot: "Your privacy is important to us. Rest assured, we take all necessary precautions to
protect your data. If you're not comfortable sharing your phone, is there another way we can
stay in touch with you?"

The chatbot should take a proactive approach, initiating interactions and adeptly transitioning
into small talk or different topics when users hesitate to share details. For example: the Chatbot
asks for the user's name, and the user declines:

User: "I'd rather not share my name."


Chatbot: "No worries, you don't have to share your name if you're not comfortable with it. So,
what brings you here today? Is there anything in particular you'd like to know or discuss?"

Once user confidence is established, the bot seamlessly returns to data collection. For example:

User: "I'm having a great day!"


Chatbot: "That's fantastic to hear! By the way, I have my birthday party coming up in a week,
and I'd like to invite you. Please share your email or phone number so I can send you the
invite?"

Moreover, this chatbot should have the capability that it can edit user data in the saved database
if the user mentions during the conversation that they provided dummy data earlier and now
wish to correct or update their details, demonstrating flexibility and trustworthiness. For example:

User: Hey when you asked me for my name I gave you a fake name as I wasn’t sure earlier, but
now since we have talked and I have gotten to know you, you can call me Rahul.
Chatbot: Ok Rahul, sure thanks for being so transparent with me, well Rahul would you like to
know more about the query you had ?..

These examples illustrate how the chatbot can engage users effectively, handle hesitations, and
transition smoothly between topics to encourage data collection indirectly.

Also, create a backend for the same to be used as a service with REST APIs for appropriate
functions/tasks the bot will need to do.

Pointers for the Solution :


● For the main conversation flow you can use the current LLMs but only the ones that are on
a similar level of GPT3 turbo. Using GPT3.5 and LLMs on its level is not allowed.
● The solution should have a function to combat prompt injection.
● Solution must include Named Entity Recognition (NER):
Implement NER to identify and extract specific pieces of information (e.g.,
names, dates, locations) from user responses. The extracted data should be stored in a
structured format. It should also have a code to check for basic validity of data for age,
name, phone number, email, etc.
● Solution must include one of the following: [Either 1 OR 2 ]
1. Sentiment Analysis: Implement sentiment analysis to understand the emotional
tone of the conversation. This can help the chatbot adapt its responses based on
the user's mood and provide more personalized interactions.
2. Multi-Language Support: Enable your chatbot to handle conversations in multiple
language inputs/outputs using Transformer models (No API-based models)

Sharing the Solution via GitHub Repository


Create a GitHub repository to host the code. Include clear instructions on how to run the code
and replicate the chat experience. Upload the code files, including the Python script(s), any
necessary dependencies, and a README explaining the project.

You might also like