Free AWS Project: Build a Chatbot with Amazon Lex in 1 hour!

[@Red-b1j]

This is quite literally one of the most adorably personable and informative tutorials I think I have ever had the pleasure of finding. I am now a subscriber for life. You are amazing. Going to be binge watching all the other AWS videos before taking my cloud practitioner and cloud architect certifications. Thank you!

[@itsnextwork]

Wow, @Red-b1j to hear such high praise is such a blessing. Thank you so so much for taking a chance on this tutorial, giving it a go and leaving the sweetest comment ever. It makes a big difference! We give nextwork.org and the community our all, and people like you keep us going 🔥 All the best with the cloud practitioner and architect certs! You might find this interesting: community.nextwork.org/c/i-have-a-question/anon-question-solutions-architect-associate

[@ahmedabdelhady4054]

This is the cutest lecture ever exists 😂, Thanks.

[@theaakashd]

0:00 😄 Helloooooo

[@itsnextwork]

Haha hellooooo Aakash!

[@itsnextwork]

"How can I get the documentation template?!" 👇Here's how! community.nextwork.org/c/i-have-a-question/documentation

[@devopsmethodology-q4x]

many many thanks Ms Natasha San...awesome ,My First AI & ML series.........next series upto... just asking...Why don’t we set the intent classification confidence score threshold to 1.0 instead of 0.4?

[@itsnextwork]

Ah what a good question! Setting it to 1.0 sounds like a dream scenario, Lex would reply with 100% certainty of the user's intent! But, that comes with it's challenges - it's actually near impossible for an AI tool like Lex (or even ChatGPT) to be 100% confidence that they understand the user's intent. If they need to wait until 100% certainty, chances are, they always trigger the fallback intent response because they're never confident enough! 🤭

[@devopsmethodology-q4x]

@@itsnextwork ohhh great.....i am just amazed ....why i never certainly know great...actually finally...solved...

[@gowfanss]

Ur a life saver.

[@itsnextwork]

Haha thank you so much! 👊

[@lancasterchrisekwueme8226]

You didn't say how to create it to production or how to integrate it in a business setting

[@gowfanss]

Can you please include two things in same project series. 1. Fallback intent 2. Getting accuracy of chatbot to 90%

[@itsnextwork]

Oooohh great idea! You should be able to update the accuracy of the chatbot by using the Intent Classification Confidence Score to 0.9 instead of 0.4. The Fallback intent we cover in future projects.

[@gowfanss]

@@itsnextwork I mean there's testing feature right where we can upload data set. There if suppose we have 500 in scope utterance and 500 fallback utterance. Then it should be able to give accurate results.

[@vishalisharani1410]

Great Natasha.

[@itsnextwork]

Wow thank you so much Vishal! Appreciate you doing this project with me.

[@javiergómez-r8g]

Buenas, el bot siempre me dice "El bot aún no está listo para pruebas" pero aún así funciona. ¿A que se debe? un saludo y muchas gracias.

[@itsnextwork]

¡Hola Javier! Es probable que este mensaje aparezca porque aún no has pulsado "Construir" en la configuración de tu bot de Amazon Lex. Asegúrate de construir el bot para que se actualice con las configuraciones más recientes y funcione correctamente. ¿Has intentado ya pulsar "Construir"?

[@TheoKol-ti5nz]

Great.

[@itsnextwork]

Thank you so much!! Good to see you here