of most developers’ work. We use tools resembling Cursor, Windsurf, OpenAI Codex, Claude Code, and so forth, to turn into way more productive at work. Nevertheless, from discussions with people working in non-technical lines of labor, I often see lots of processes that will be optimized using AI.
In this text, I’ll show how other non-tech industries can profit from using AI. I’ll highlight this by automating the technique of finding prospects with AI, a typical sales job that requires lots of manual work. The goal of the article is to focus on how even non-technical people could make use of the most recent coding agents to create powerful automation tools.
Throughout the article, I’ll be highlighting my essential suggestions with:
It is a essential tip
Why do we’d like automatic prospect finding
Business development representatives work with:
- Finding interesting prospects
- Retrieving their contact information, resembling their LinkedIn profile or email address
- Reaching out to prospects, attempting to establish a gathering
From there, an account executive normally takes over, though I’ll be specializing in how one can optimize the primary three steps.
This three-step process is generally quite extensive, as finding prospects online requires scouring through lots of LinkedIn profiles or other web sites to seek out interesting firms. After finding an organization, you usually start in search of specific people within the organization to achieve out to. It will normally be key decision makers, who is usually a middle manager in a bigger company, or a CFO in a smaller company. After finding the right person, you could get their contact information, normally found on LinkedIn or on the corporate’s website. Finally, you need to reach out to this person with a personalised message.
Finding interesting prospects
I’ll start developing this tool using Claude Code. You possibly can essentially use any coding tool there, resembling Codex, Cursor, Windsurf, Replit, and so forth. The essential point is the commands you employ to make the applying.
All the time use planning mode before making a recent application
I all the time start off using , which tells the model to read your prompt, create a step-by-step plan, and ask any clarifying questions. This may be very useful, because it helps you cut down the scope of your application and ensures you provide your thoughts on any ambiguity. For instance, I got prompted on:
- Which location are you targeting (in my case, I’m taking a look at prospects in Norway)
- Which programming language do you would like your application in (I selected Python, but TypeScript or other languages are in fact also possible)
- What are some web sites you should have a look at (in Norway, has lots of information on firms and employees, so yow will discover similar tools, and even ask your coding agent to seek out these web sites itself using an online search)
- What output format do I prefer (I selected an Excel sheet)
These are all great inquiries to make clear, which is why using is so useful.
Also, ensure that to prompt your model to make use of publicly available API’s, and only search for information from relevant firms. Fetching personal information might be against the regulations, depending on where you reside.
All the time provide your coding agent as much tools as possible. It might be MCP servers, OpenAI credentials, or access to any programs through an API.
Moreover, I also made an OpenAI key, which I told Claude Code it could load from my .env file with OPENAI_API_KEY. This is helpful because lots of operations will probably require using an LLM to seek out or process information. Thus, providing Claude Code access to a robust API service may be very vital. If you have got other relevant API’s, you need to also provide Claude Code with the documentation for these API’s and tell it to make use of them.
I, for instance, informed Claude that it has access to web search, and may perform it with the function below:
response = client.responses.create(
model="gpt-5",
tools=[{"type": "web_search"}],
input="What was a positive news story from today?"
)After I answered all of the clarifying questions, I told Claude to begin constructing, and it built an application to seek out an inventory of prospects, returned in CSV format. This covers each step 1 and a pair of, of first finding interesting prospects, and getting a hold of their contact information.
After finding all of those prospects, you need to also do a manual review, ensuring correctness. Moreover, I like to recommend prompting GPT-5 or an equivalent model to undergo your results and confirm any inconsistencies.
Have an LLM undergo your results to confirm correctness
Lastly, it’s also vital to abide by regulations when finding prospects. It is best to only be finding relevant firms online, after which finding individuals manually to abide by GDPR regulations. Thus, to seek out additional information, resembling the name, email, and role of individual prospects, I manually find the data from the businesses supplied with my application.
Reaching out
After finding contact information, you now need to achieve out. You possibly can read lots of stats and knowledge on how one can perform cold emails, but I’ll not get into that here, since my focus is on the technology and the way we are able to use it to optimize our processes.
Up so far, I assume you’ve fetched an inventory of relevant prospects, including their contact information, and also you’re ready to begin reaching out. We now wish to create customized messages to every individual, which is luckily a task that LLMs are really good at.
LLMs are in a position to craft personalized messages
For instance, at this stage, I even have the next information per prospect:
- Individual name
- Individual email
- Individual role
- Company name
- Company size
- Company revenue
Also, usually, adding more information to your prompts is generally higher. In case you prefer a selected style or tone in your emails, you need to add that information. idea can be to indicate examples of you’re previous emails, highlighting the way you write emails yourself, and thus using few-shot prompting to enhance output quality.
Add as much information to your prompts as possible
I’ll now use this information to draft up a customized message. This will be done relatively simply using GPT-5 for instance:
prompt = f"""
You're an authority at creating personalized emails. You're given information
about a person and need to create an email to achieve out to them for the
first time.
Name: {name}
Email: {mail}
Role: {role}
Company name: {company_name}
Company size: (company_size}
Company revenue: {company_revenue}
Create each a subject tag, and the complete email, including no other comments
or reasoning.
"""
client = OpenAI(api_key=OPENAI_API_KEY, base_url=API_URL)
import os
from openai import OpenAI
result = client.responses.create(
model="gpt-5",
input=prompt,
reasoning={ "effort": "low"},
text={ "verbosity": "low" },
)When you have got the outline of the message, you possibly can then manually tweak and optimize it for the person you’re reaching out to, given the data you’ve found.
You possibly can now reach out using these emails. To avoid breaching any terms of service and spamming people, I like to recommend reaching out manually and never using an automatic service. I don’t condone spam emails or similar. AI is barely used to show you how to and make processes more practical, to not remove all humans from the loop necessarily.
Conclusion
In this text, I’ve highlighted how you should utilize the most recent coding tools, resembling Claude Code, to automate some processes. In this text specifically, I covered how one can optimize the technique of reaching out to prospects by utilizing AI to routinely find relevant firms and routinely create emails while staying compliant. We’ve seen immense progress with AI in the previous couple of years, but I still consider AI is lagging behind on the implementation side. Thus, in case you will be fast at integrating AI into your day by day life, you possibly can have an enormous advantage over your peers.
👉 Find me on socials:
🧑💻 Get in contact
✍️ Medium
You too can read my other articles:
