Job tracker and calls to OpenAI

API
1

I have a CSV file of IT companies.
I want to use OpenAI API to automate the calls to the chat about each company in the entry.

How can I use it for this?

I want:

  • for every entry of companies, OpenAI API will make a search, and return the category or industry it is aimed at
  • for every entry … , OpenAI API will browse online for new jobs, and return a list of the jobs and links to them, people who work in the company, their LinkedIn pages, etc

Any idea how to get started with it?

2

Some models do support websearch, but it can be very expensive depending on the usage (like a big list of companies…).

1 Like
3

You’re right, for sure as per the second goal.

But maybe for the first goal, I could send a single websearch with all of the companies list?

I’m not sure if using websearch is even the solution I need.

4

You are charged per tokens (it might perform many multiple calls even with one request, multiplying the costs), also I’m not sure a single search would be successful. But you can try.

1 Like
5

The web search function that OpenAI offers isn’t good for increasing AI knowledge. It is good for dumping out results in a chatbot session like a search engine.

Here’s how you can automate your way through a CSV, making one API call per company that is extracted (supposing there is a unique column having “name”), and asking the AI about it from built-in knowledge. This is Python code.

import csv
import os
import time
import httpx

COMPANIES_CSV = "companies.csv"
OUTPUT_CSV = "company_industry.csv"
NAME_FIELD = "name"  # Change this if your column is named differently

def find_name_field(header):
    # Try to find a column containing 'name' (case-insensitive)
    for col in header:
        if "name" in col.lower():
            return col
    raise ValueError("No column containing 'name' found in CSV header.")

def ask_industry(company_name):
    prompt = f"In one sentence, as your only response, what is the main industry of this company? {company_name}"
    payload = {
        "model": "gpt-4o",
        "input": [
            {
                "role": "user",
                "content": [
                    {"type": "input_text", "text": prompt}
                ]
            }
        ],
        "tools": [],
        "text": {"format": {"type": "text"}},
        "temperature": 1,
        "top_p": 0.05,
        "reasoning": {},
        "stream": False,
        "max_output_tokens": 128,
        "store": True
    }
    response = httpx.post(
        "https://api.openai.com/v1/responses",
        json=payload,
        headers={
            "Authorization": f"Bearer {os.environ['OPENAI_API_KEY']}",
            "OpenAI-Beta": "responses=v1"
        },
        timeout=60
    )
    response.raise_for_status()
    for output in response.json().get("output", []):
        for content in output.get("content", []):
            if content.get("type") == "output_text" and "text" in content:
                return content["text"]
    return ""

with open(COMPANIES_CSV, newline='', encoding="utf-8") as infile:
    reader = csv.DictReader(infile)
    if NAME_FIELD not in reader.fieldnames:
        NAME_FIELD = find_name_field(reader.fieldnames)
    companies = [row[NAME_FIELD] for row in reader if row.get(NAME_FIELD)]

with open(OUTPUT_CSV, "w", newline='', encoding="utf-8") as outfile:
    writer = csv.DictWriter(outfile, fieldnames=["name", "answer"])
    writer.writeheader()
    for company in companies:
        print(f"asking AI about {company}...")
        answer = ask_industry(company)
        writer.writerow({"name": company, "answer": answer})
        time.sleep(5)

This example script shows how to process a CSV file containing company data and use an AI API to find out the main industry for each company. Here’s how it works:

  1. CSV Reading:
    The script reads a file called companies.csv and looks for a column with “name” in its header (you can change the column name if needed).

  2. API Calls:
    For each company name, it sends a question to the AI:
    “In one sentence, as your only response, what is the main industry of this company? {company name}”

  3. Progress Reporting:
    As it works, the script prints which company it’s asking about.

  4. Rate Limiting:
    It waits 5 seconds between each API call to avoid sending too many requests too quickly.

  5. Results Saving:
    The answers are saved to a new CSV file called company_industry.csv, with columns for the company name and the AI’s answer.

You can change the NAME_FIELD variable at the top if your CSV uses a different column name for company names. Make sure you have your OpenAI API key set in the OPENAI_API_KEY environment variable before running the script.

2 Likes