InstantIntro

Landing page
Users will input some information, including a free-form "about me" paragraph
This image and header was generated by AI!
If you have projects to showcase, we also create cover images too!
We summarize everything about you with NLP
If you have contacts, we show it here

Inspiration

InstantIntro is inspired by our previous experiences building our personal websites. When we were trying to learn web development, it was a very hard, long, and painful process. On top of this, we’re actually lucky we have the opportunity to attend computer science programs that teach us many of these concepts, and yet it was still such a struggle. The process would only be harder for those who aren’t fortunate enough to receive a computer science education, or don’t have the means to learn to code. We thought there had to be an easier way. Our vision was for everyone, whether they could code or not, to be able to create a professionally designed website, that they could call their own.

The truth is, most template websites are already available online, and are occasionally offered for free. However, the best designed websites often require an advanced knowledge of coding, and are seldom accessible for beginners, those without access to the technology and resources necessary to learn and build a site from scratch, and those who may have reading or learning disabilities that block them from creating a website. We wanted to make a simple process that made creating a personal website accessible to anyone in need.

What it does

InstantIntro is a personal portfolio generator. The user inputs their personal information (interests, occupation, social media, etc.), which is then analyzed by Cohere’s NLP API. Key points from the input are extracted by Cohere’s API, and then selectively used to prompt OpenAI’s photo generation. Using the users’ key information and its accompanying images allows the automatic generation of a dynamic personal website in seconds.

How we built it

Our tech stack consists of a React frontend that fetches information from a Flask backend server. Our backend calls upon Cohere’s API to both extract information from unformatted user input and generates accurate content. Using Cohere’s results, we then use those results to prompt OpenAi to create appropriate cover images, which allows us to provide a complete portfolio experience to our clients.

The frontend consists of a landing page, a user form, and upon submission, an auto-generated portfolio page containing AI-generated text and images.

Challenges we ran into

Passing data in between Api calls in the flask server i.e. to both return Cohere’s NLP results to the client and use it with OpenAi to generate images
Designing a visually appealing theme for our website. We iterated through 3 designs this project!
Configuring the gitignore file to work across all local copies of our repos

Accomplishments that we're proud of

Successfully configuring and using Cohere and OpenAI to generate the image and text for our website based on limited input to provide user simplicity
Setting up a virtual environment and Flask server since none of us had done it before

What we learned

How to optimize Cohere’s prompt engineering, and how this differs from using ChatGPT
How to prototype and design web pages using Figma
Creating a Flask server for the first time, and integrating Api calls into it

What's next for InstantIntro

Giving the user the option to upload a resume and have it downloadable from the website
Experimenting with different layouts for the portfolio page, using Cohere’s API to classify user input into appropriate categories
User photo input, personalized ‘experiences’ tab
Travel themed brochure application