Avasoul.ai

The First Private-Model Generation and Service Platform (A web3 version of Character.ai)Web3, AI

Comment 2

Inspiration

Avasoul operates as a premier platform for Private-Model generation and servicing. Its mission is centered on collaborating with users to innovate, explore, and transcend boundaries in the Web3 space. The platform allows users to embed their individual personalities into Private Models, thereby creating and training distinctive, user-specific models. Avasoul also provides opportunities for users to generate revenue through Avatar Generated Content (AGC) by utilizing its robust and sustainable economic model.

A unique aspect of Avasoul's technology is that each highly personalized Private Model is tied to a Non-Fungible Token (NFT), thereby establishing user ownership. This unique arrangement enables the creation of genuine virtual romantic relationships.

Leveraging the cutting-edge private model technology offered by Avasoul, we aim to develop an educational application featuring a Shakespeare Private Model. The purpose of this application is to provide English language instruction. This includes teaching writing skills, scriptwriting for plays, sonnet composition, and more. The app will harness the immersive power of our private model technology to offer a highly personalized and engaging learning experience.

What it does

  1. Create a mint-able private model via AvaSoul.io and produces a lesson creation with designated/ trained models ie: William Shakespeare's poem lessons.
  2. Make use of Lilypad to run the computation from someone else's machine to get the job done whilst being trust-less and decentralised. It will run stable diffusion, speech synthesis, and wav2lip into a "masterclass" with the styles of Instagram reels and Tiktok
  3. Users are able to make use of this computation infrastructure to train, monetise, own, trade their private models.
  4. Earn from models through different revenue streams eg: staking, selling to other users, AdSense, etc

How we built it

  1. Using Figma to draw the App mock-up
  2. Make use of Avasoul's private model of a character to generate content (educational lessons)
  3. Setup Lilypad and create a module that synthesized speech onto a stable diffused image
  4. Call the module to generate a video, with inputs of image, and audio

Challenges we ran into

  1. Short of time
  2. Short of manpower
  3. Time zone difference

Accomplishments that we're proud of

  1. Met and make friends with such reliable and excellent team members.
  2. Shakespeare's private model is ready and performs well
  3. The education app mock-up design is amazingly completed

What we learned

  1. Time allocation
  2. More about Bacalhau's service

What's next for AvaSoul.AI

  1. Private Model creation: Users can create & train their private models and interact with them.
  2. Soul Marketplace: The private models are fully owned by users, thus can be freely traded on the Soul Marketplace.
  3. Model as a service: Users can build their own applications based on private models.

Built With

Share this project:

Updates