Transform your decision-making journey into an exhilarating adventure responsibly with AIGuru – the Guru of All Creative Assistants!

Inspiration

In the domain of AI, decision-making is a crucial part. Many times, we need to choose between 2 different options, some examples are

  • Shall I eat a burger or a pizza? 🍔 🍕
  • Shall I go to the gym or play a game? 🏋️ ⚽️
  • Shall I buy a car or a bike? 🚗 🏍️
  • Shall I invest in stocks or mutual funds? 📉 💵
  • Shall I go for a movie or a party? 🎬 🍾

and many more small and big decisions. The AIGuru 🤖 is such a tool that helps in making decisions by providing the best possible solution to a problem along with suited explanation. No matter how important or trivial the decision is, AIGuru 🤖 will help you make the best decision, responsibly.

Our Key Inspiration was also to utilize AWS Partyrock to implement a practical aspect of Responsible AI, for which we drew motivation from following research paper 📝

What it does

  • AIGuru 🤖 analyses the User's Input, which usually are just 2 options to choose from, seperated by ;
  • Then it asks user to provide answer to follow-up questions ❓ that it generates, by understanding options, in a few words.
  • User can answer as many questions as they want, in like a word or a phrase, and when they are done, they can ask AIGuruChat the same options. 💬
  • AIGuruChat will then provide the best possible solution to the problem along with suited explanation 🤔, that it draws from answers and follow-up questions.
  • Finally, AiGuruImageGenerator provides a visual representation of the decision made by AIGuruChat 🌄, to motivate user to take action.
  • Moreover, AIGuruChat, can give you answers and explanations in multiple languages ⽂, as it is powered by Claude Model

How we built it

  • We used AWS Partyrock to implement AIGuru, AIGuruChat and AIGuruImageGenerator 💻
  • We used Claude Model to provide answers and explanations in multiple languages 🧠
  • We used StableDiffusion XL for Image Generation 🌄

The benefit of AWS Partyrock are,

  • It provides a no-code interface to deploy powerful AI models provided by AWS Bedrock and 3rd party model providers
  • Give it the problem statement, and AWS Partyrock, provides necessary widgets and prompt boilerplates to start with
  • It provides with necessary UI components to interact with AI models
  • It also provides with a sharable Link to share and showcase project, thus making Generative AI, accessible to everyone

Challenges we ran into

  • We had to learn how to use AWS Partyrock, as it was our first time using it 💻
  • We had to tune models a bit (temperature & top-p), but thanks to AWS Partyrock, it was easy to do so 🛠️
  • We had to write proper prompts, to get desired answers from AI models 💬

Accomplishments that we're proud of

  • We are proud of the fact that we were able to implement a practical aspect of Responsible AI, by utilizing AWS Partyrock ⛑️
  • We are proud of the fact that we were able to provide a tool that helps in making decisions by providing the best possible solution to a problem along with suited explanation ❓
  • We are proud of the fact that we were able to provide a tool that can give you answers and explanations in multiple languages and generate visual representation of the decision made ⽂
  • We are proud to do all this without writing a single line of code, thanks to AWS Partyrock 📜

What we learned

  • We learned how to use AWS Partyrock 💻
  • We learned about Responsible AI ⛑️
  • We learned about Generative AI 👾
  • We learned about AWS Bedrock and 3rd party model providers

What's next for AI Guru

  • We plan to introduce more models to analyze answers to AIGuru follow-up questions, and cumulatively provide the best possible solution to the problem along with suited explanation
  • We plan to introduce more language support to AIGuruChat
  • Sometimes, AIGuruImageGenerator, changes visual representation based on AIGuruChat's language, we plan to fix that
Share this project:

Updates

Submission history