Vangaurd

• 

  Dashboard example

Inspiration

while majority of candidates target narrower E-commerce use case and treat call-E as a universal LLm source i made a website that can take any LLM/AI model with call-E and tracking tokens,performance and confidence scores extensively with charts !

What it does

Adds the ability to users to add any LLM/AI model (even locally run) and use , question and prompt it side by side with call-E resulting in large scale real world data fetching and extensive planing and discussion with LLM before prompting call-E

How we built it

i used Typescript as my main language with css for some styling and react and vike for the framework , i used express, zod and Typescript an node for the backend with vercel,s ai SDK and call-E SDK

Challenges we ran into

makeing and unifying any LLM compatibility and making responsive dashboards with privacy first platform

Accomplishments that we're proud of

it is compatible with any AI/LLM model and uses responsive dashboard with user privacy sights and targeting a rather future capable Idea , it can be extended INSANELY yet staying flexible to each user

What we learned

How to make production levels applications with typescript and zod for stricit schemeing results and inputs to backend pipelines

What's next for Vangaurd

it,s main idea and foundation now is extremely solid witht he extensible idea of extending the use of call-E with LLM/AI and then in future makeing whole AI orchestrated real world data with infinite possibiities

Built With

• ai
• fullstack
• llm
• react
• typescript