Turn 1 Tech Skill Set Into 5 Income Streams: The Software Engineer Edition (All 5 Prompts)

You came here from Instagram, so let's skip the warm-up. Below is the five-stream framework run on one title, the software engineer, with a price, a buyer, a first move, and the Claude prompt for each one. Copy what you need.

If you want the framework itself, it's in the original post. If you want to see it run on a job that has nothing to do with code, there's the nursing version. This one is for engineers.

The premise

Your salary is one way to sell your expertise. Not the only way.

Someone already pays you to solve technical problems. Now ask who else needs that skill. The answer is: a lot of people who will never post a job listing for it.

One skill set. Five possible offers:

1. The service. Someone pays you to do it.
2. The teach. Someone pays you to explain it.
3. The product. Someone pays once for the version you already made.
4. The tool. Someone pays monthly for the thing you built.
5. The audit. Someone pays for your judgment on their work.

No new degree required. Let's run it on one title.

Income stream 1: The service

AI support assistant

What you'd do. Build a small business a chatbot that answers customer FAQs, books appointments, and hands tricky questions to a human.

Price. $3,000 fixed.

First move. Build a demo for one business. Not a pitch deck. A working demo with their actual FAQs loaded in.

The pitch. Sell "fewer missed customers." Not "I know LLM APIs." The owner of a dental office does not care what model you used. They care that the phone stops ringing at 9pm with questions a bot could answer.

The prompt:

I'm a software engineer who works in [stack]. Ask me about the systems I've built and the problems they solved.

Then suggest 3 types of small businesses that would pay for an AI support assistant, a fixed-scope package with clear limits, and a short DM asking about their current customer questions.

Don't invent results.

That last line matters. Without it you get a DM full of made-up case studies. With it you get a question you can actually send.

Income stream 2: The teach

Build your first AI agent

What you'd do. A self-paced mini-course where developers build one working AI agent from start to finish.

Price. $139.

First move. Post one free lesson. If people finish it and ask what's next, you have a course. If they don't, you saved yourself three months.

The rule. Teach one project they can finish. Not your entire computer science education. The buyer is a developer who wants to ship an agent this weekend, not someone who wants to understand transformers.

The prompt:

Act as a curriculum designer for self-paced developer courses. Interview me one question at a time about how I'd build [agent] for a developer who has never done it, and where beginners usually get stuck.

Then, using only what I've told you, outline 3 short lessons with one clear outcome each, a final project they can finish in a weekend, and a troubleshooting exercise based on the mistakes I described.

Flag any code I need to test, and write a one-line course promise.

"Using only what I've told you" is what keeps this from turning into a generic AI course outline that already exists in forty places.

Income stream 3: The product

The SaaS starter kit

What you'd do. Package the setup you rebuild on every project (auth, database, payments, email, the works) into code someone can buy and launch on.

Price. $199 per kit.

First move. Check what you've built twice. If you have set up the same auth flow on two side projects, that is the product. You already built it. You just never priced it.

One thing to get right. This has to be code you built on your own time and own outright. Not something lifted from your employer's repo. The prompt below assumes that, and so should you.

The prompt:

Act as a product strategist for developer tools. Here's a starter kit I built on my own time and own outright: [description].

Interview me one question at a time about who it's for, what it saves them, and what still needs testing.

Then write a product name, a one-line promise, a setup guide outline, and a launch checklist. Flag any licenses or dependencies I need to review before selling.

The license flag is not optional. Some packages you depend on cannot be resold inside a commercial kit, and you want to find that out before the first customer does.

Income stream 4: The tool

The AI receptionist

What you'd build. An AI voice agent for salons that answers calls and books appointments 24/7.

Price. $49 a month.

First move. Free pilot with one salon. One. Not a landing page with a waitlist. A real salon with real calls so you find out what breaks before you charge anyone.

The pitch. Every missed call is a missed client. Your tool never misses one. That sentence is the whole sales page.

The prompt:

I'm a software engineer who works in [stack]. Ask me what I know about voice APIs, scheduling integrations, and the businesses I can reach.

Then spec an AI receptionist for one niche: the calls it handles, when it hands off to a human, the build order, and estimated cost per call.

Write 5 questions to test whether owners would pay monthly. Flag anything I'd need to verify about call recording consent.

Cost per call is the number that decides whether $49 a month is a business or a hobby. Make the prompt do that math before you write a line of code.

Income stream 5: The audit

The vibe-code audit

What you'd do. Find the bugs and security holes in AI-built apps by non-technical founders before they launch.

Price. $499 per audit.

First move. A sample teardown. Pick one public vibe-coded app, write up what you found, and post it. The teardown is the portfolio.

Why this works. They can generate code. You know when it's wrong. There are more founders shipping AI-generated apps right now than there are engineers willing to look at them, and most of those apps have an exposed API key somewhere.

The prompt:

Act as a client-acquisition strategist for technical consultants. Interview me one question at a time about the most common bugs and security issues I find in code reviews.

Turn my answers into a 10-point audit checklist for AI-built apps. Then suggest 3 places to find non-technical founders and write a short message offering a fixed-scope review.

Don't claim their app has problems I haven't found.

That last line is the difference between a consultant and a spammer. You do not tell a founder their app is broken. You offer to look.

Which one to start with

Same rule as always: pick by speed to first dollar, not by how it sounds in your bio.

For an engineer, that is the service. One business, one demo, one invoice. You already know how to build it, and the buyer already has a budget for the problem.

Make the audit second. The same ten-point checklist resells to every founder you find, and the sample teardown you post is doing your marketing for you.

The tool comes last, even though it is the one you want to build first. Tools need users. Services need one email. Build the receptionist once a salon owner from your audit work asks you to.

You already learned how to build

Now learn another way to sell what you know. The code was never the bottleneck. The buyer count was.

Swap the stack and every prompt above still works. Run it on your own title next.

I'm Naya, a self-taught engineer building things with AI in public. If this was useful, the newsletter goes deeper every week: Disgustingly Ambitious.