AI-driven Industry, Market, Value Proposition, and Target Audience Recognition

Most companies try to stay ahead of the curve when it comes to visual design, but for Planetaria we needed to create a brand that would still inspire us 100 years from now when humanity has spread across our entire solar system.

I knew that to get it right I was going to have to replicate the viewing conditions of someone from the future, so I grabbed my space helmet from the closet, created a new Figma document, and got to work.

Project Challenge

In my mission was to gain a more comprehensive understanding of companies in various sectors, I have faced with a significant challenge: the recognition of industry, market, value proposition, and target audience. While traditionally such analysis would require extensive manual research, my goal was to automate this process using AI, ensuring high levels of accuracy, efficiency, and scalability.

Design Approach

My design approach revolved around automating data extraction from company websites and processing this information through the OpenAI API to identify key business attributes.

The journey began with extracting the HTML code from a company's website. From this code, I then isolated the textual content and summarized it to create a concise yet comprehensive overview of the company. This summary was then passed through the OpenAI API, which was programmed to identify the company's industry, market, value proposition, and target audience.

The Solution

By harnessing the power of the OpenAI API, I was able to develop an AI-driven solution that automates the recognition of a company's key attributes.

import requests
from openai import OpenAI
from bs4 import BeautifulSoup

company_url_list = ['https://tesla.com', 'https://amazon.com', 'https://apple.com']

# Get the HTML code for each company's website
for company_website_url in company_url_list:
    response = requests.get(company_website_url)
    html_code = response.text

    # Extract textual content from HTML
    soup = BeautifulSoup(html_code, 'html.parser')
    text_content = ' '.join([p.text for p in soup.find_all('p')])

# Call the OpenAI API with text_content using
client = OpenAI()

response = client.completions.create(
  model='gpt-3.5-turbo-instruct',
  prompt=f'Get the needed JSON information from the provided HTML:{text_content}\n\nReturn the values in JSON format: {"company_name": "", "industry":"", "market":"", "value_proposition":"", "target_audience":""}',
  temperature=0,
  top_p=1,
  frequency_penalty=0,
  presence_penalty=0
)

# Process and display the results
json_response = response.choices[0].text

The results of this process are astonishing. With our AI-driven solution, I have been able to automate a process that would traditionally require hours of manual labor. Moreover, the accuracy of the information obtained is exceptionally high, thanks to the advanced capabilities of the OpenAI API.