How to transform workflows with AI

of the work of many developers. We use tools like pointer, Windsurf, OpenCEICEX, Claude code, and more, to be more productive at work. However, in conversations with people who work in non-technical fields, I often see many processes that are not optimized using AI.

In this article, I will show how some non-tech industries can benefit from using AI. I'll highlight this by making the process of finding prospects with AI, a common sales activity that requires a lot of manual work. The goal of the article is to highlight how non-technical people can use coding agents to create effective automation tools.

Throughout the article, I'll be highlighting my top tips for:

This is the main tip

Why do we need automatic discovery you get

Business development representatives work with:

1. Finding Exciting Prospects
2. Retrieving their contact information, such as their LinkedIn profile or email address
3. Reaching out to prospects, trying to set up a meeting

From there, the account manager usually takes over, although I will be focusing on how to do the first three steps.

This three-step process is very common, since finding prospects online requires hitting many LinkedIn profiles or other websites to find interesting companies. After you found a company, you usually start looking for people in the organization to reach out to it. These are often top decision makers, who may be a middle manager in a large company, or a CFO in a small company. After finding the right person, you need to find their contact information, usually found on LinkedIn or the company's website. Finally, you have to reach this person with a personalized message.

Finding Exciting Prospects

I will start developing this tool using Claude's code. You can use any encoding tool there, such as codex, pointer, Windsurf, repeater, etc. The main point is the instructions you use to make the request.

Always use edit mode before creating a new program

I always use it first Editing modewhich tells the model to learn at your own pace, create a step-by-step plan, and ask clarifying questions. This is very useful, as it helps you to reduce the scale of your application and ensure that you give your thoughts to any flexibility. For example, I was conceived in:

• Where are you targeting (in my case, I'm looking for prospects in Norway)
• What programming language do you want your app to be in (I chose Python, but books or other languages ​​are possible)
• What other websites do you want to view (Norway, Prof.no You have a lot of information on companies and employees, so you can find similar tools, or ask your coding agent to find these websites themselves using a web search)
• What is my preferred output format (I chose bold sheet)

These are good clarifying questions, which is why you use them Editing mode it is very useful.

Also, make sure you prompt your model to publicly use the API, and only request information from the right companies. Downloading personal information can have regulations, depending on where you live.

Always provide your agent with as many tool codes as possible. It can be MCP servers, Opelai authentication, or access to any programs via API.

In addition, I also made an Opelai key, which I told Claude Code could load from my .Env file with OpenAi_API_KEY. This is useful because many jobs will likely require using an LLM to obtain or process. Therefore, providing Claude's code access to a powerful API service is very important. If you have other relevant APIs, you should also provide Claude with the code and documentation for these APIs and tell it to use them.

I, for example, know from Claude that it has access to web search, and can do 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 the clarifying questions, I told Claude to start building, and build an application to get a list of prospects, which came back in CSV format. This includes both steps 1 and 2, starting with finding interesting prospects, and getting their contact information.

After finding all these prospects, you should also do a manual review, to ensure accuracy. In addition, I recommend promoting GPT-5 or a similar model to go through your results and confirm any inconsistencies.

LLM has gone through your results to ensure accuracy

Finally, it's also important to follow the rules when you get prospects. You should only find the right companies online, then find people manually to comply with the GDPR rules. Therefore, to obtain additional information, such as the name, email and role of individual prospects, I receive information from the companies that provided my request.

External access

After getting the contact information, now you need to access. You can read a lot of statistics and details on how to do cold emails, but I will not go into this, because my focus is on technology and how we can use it to improve our processes.

By now, I'm assuming you've downloaded a list of relevant prospects, including their contact information, and you're ready to start reaching out. Now we want to create customized messages for each person, which is fortunately a job that LLMS does very well.

LLMS can create customized message boats

For example, in this section, I have the following information for each prospect:

• Each word
• Individual email
• Each role
• Company Name
• Company size
• Company money

And, in general, the more details you add to your ProFeps the better. If you prefer a certain style or tone to your emails, you should add that information. It's also a good idea to show examples of previous emails, to highlight how you write emails yourself, and thus use a few clicks to improve the quality of the output.

Add as much detail to your motivation as possible

Now I will use this information to write a customized message. This can be done by simply using GPT-5 for example:

prompt = f"""
You are an expert at creating personalized emails. You are given information
about an individual and have to create an email to reach 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 both a subject tag, and the full 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 an outline of the message, you can then go ahead and prepare it for the recipient, given the information you have received.

Now you can reach using these emails. To avoid violating any terms of service and spamming people, I recommend manual access and don't use the automated service. I do not allow spam or similar emails. AI is only used to help you and make processes more efficient, not to actually remove all people from the loop.

Lasting

In this article, I've highlighted how you can use the latest coding tools, such as Code Claude, to change some processes. In this article specifically, I covered how to optimize the prospecting process by using AI to automatically find the right companies and automatically create emails while staying compliant. We've seen a lot of progress with AI over the past few years, but I still believe that AI is largely behind the price point. Therefore, if you can quickly integrate AI into your daily life, you can have a huge advantage over your peers.

👉 Find me in the community:

📩 Subscribe to my newsletter

🧑💻 Get in touch

🔗 lickEdin

🐦 X / Twitter

✍️ Medium

You can read my other articles: