Apollo Ai

Apollo Ai: Intelligent Conversational Voice Assistant, Powered by AI.

Comment

Inspiration

Our project was inspired by a heartwarming story of a visually impaired student and her mother, who received an honorary degree in recognition of the mother's dedication to helping her daughter complete her law degree. This story highlighted the absence of assistive technology and how, with the right support, people can overcome significant challenges. We were motivated to build Apollo AI as a voice assistant that could make everyday tasks easier for everyone, leveraging advanced AI to provide more natural, intelligent support for students, much like how the student's mother provided continuous assistance throughout her daughter’s studies.

What it does

Apollo AI is a voice assistant powered by Azure’s OpenAI GPT-4o mini model. It handles complex queries and conversations, offering smart responses beyond basic Alexa voice commands. Apollo AI can process text, speech, and even images, aiming to assist users in real time with natural, intelligent interactions.

How we built it

We developed Apollo AI using Alexa Skills Kit on an Alexa Echo Dot 3 for voice interaction and integrated it with Azure OpenAI's API to process user input. We also created a custom 3D-printed housing for the Echo Dot to give the assistant a personalized, sleek design representative of Apollo Ai's innovative nature. This blend of AI and physical design helped create a unique and highly functional voice assistant.

Challenges we ran into

  • Latency from API calls affected the smoothness of conversations, so we ended up using GPT-4o-mini.
  • Alexa's limitations posed some difficulties in terms of customization and flexibility, namely the time display and possible voices.
  • Multiple 3D printing possibilities were explored to ensure the Echo Dot fits perfectly within the custom housing.

Accomplishments that we're proud of

We successfully integrated GPT-4o-mini into an Alexa skill, elevating its conversational abilities. Additionally, the custom-designed Echo Dot housing added a unique element, turning a functional device into a personalized, aesthetic piece of tech.

What we learned

We gained experience in natural language processing, API integration, and 3D design. We also learned how to optimize voice interactions using AI models and handle real-world design constraints with physical devices.

What's next for Apollo AI

We plan to expand Apollo AI’s capabilities to process images and files fully. We also aim to reduce latency in API calls for faster, more seamless interactions, and to explore integrating the assistant into other platforms.

Built With

Share this project:

Updates