Ideation

The genesis of this project was a comprehensive dialogue with sponsors from Constellation Brands regarding potential use cases that would enhance their organizational workflows which supplemented the idea of creating a platform that would augment productivity. On that account, I undertook the task of developing a virtual assistant pipeline with a primary emphasis on reliable information retrieval, minimizing the need to navigate through extensive data repositories and conduct exhaustive web searches for swift answers. Additionally, to further streamline the process and for personal convenience, I decided to integrate Siri functionality to eliminate the need for manual text input when posing queries.

What it does

The choice of Athena, named after the goddess of wisdom, war, and crafts, was aptly suited to the objectives of this implementation. This iOS application facilitates swift access to a broad spectrum of documents such as policy manuals, thereby aiding users in expediting their manual tasks. Moreover, given the paramount significance of marketing in business, an AI image generator has been incorporated to enable individuals to unleash their creativity and transform their ideas into compelling visual art. Lastly, the Siri integration enhances accessibility by enabling users to engage with the AI through voice commands.

Additional Features

  • Save chat history for FAQs
  • Option to save or discard conversations
  • GDPR-compliant document storage

Technologies Used

  • OpenAISwift wrapper
  • SwiftUI

Challenges I ran into

One of the greatest challenges with the latest advancement of AI Technology is exploring areas that will help member engagement and increase productivity while conserving security and authenticity. Coming up with a system architecture and a data pipeline to achieve this was extremely challenging.

Accomplishments

  • Bulletproof accuracy
  • Fast response times
  • Voice activated interface
  • Easy to use UI
  • Enterprise-ready stack

Learning

  • Informational trustworthiness cannot be compromised
  • Recognizing success involves identifying the value of use cases

What's next for Athena AI

  • Async database queries
  • Load testing
  • Add clickable links to UI
  • User option to manually load documents

Built With

Share this project:

Updates