How to Get Your First AI Agent Result in 3 Minutes (Step-by-Step)

You signed up for an AI agent tool. You're excited about the possibilities. And then... nothing. You stared at the dashboard, unsure what to do next.

This happens to almost everyone. The gap between "I have an account" and "I just got real work done with AI" is where most users give up.

This guide closes that gap. In the next 3 minutes, you'll go from zero to a completed AI agent task with real output you can use right now.

What you'll produce: A competitor analysis report for your industry.

Related: 10 AI Agent Workflows You Can Set Up Today · How to Build an AI Research Team in 5 Minutes · Free AI Agent Tools Guide

The 3 Things You Need (2 Minutes to Prepare)

1. An API Key ($5 minimum)

AI agents need access to a language model. You get this through an API key from a provider:

• Anthropic (Claude) — Best for research and analysis. $5 in credits covers 100+ tasks.
• OpenAI (GPT-4) — Strong for creative tasks. Similar pricing.

Go to either link, create an account, add $5 in credits, and generate an API key. This takes about 90 seconds.

2. A Free Ivern Account (30 seconds)

Go to ivern.ai/signup. Sign up with email or Google. No credit card needed.

3. Your API Key Connected to Ivern (30 seconds)

In Ivern, go to Settings → Connections. Click Add API Key. Paste your Anthropic or OpenAI key. Click save.

Your key is encrypted immediately. Ivern uses your key directly — zero markup on AI costs.

That's it. You're ready to run your first task.

Your First Task: Competitor Research Report (60 Seconds)

Step 1: Create a Squad

From the Ivern dashboard, click Create Squad. Name it "Research Team."

Add these two agents:

Agent	Role Template	Purpose
Researcher	Researcher	Gathers information about each competitor
Writer	Writer	Compiles findings into a polished report

Click Create Squad.

Step 2: Run This Exact Prompt

Go to your squad's task board. Click New Task. Paste this:

Analyze the top 3 competitors in [your industry here]. For each competitor, identify their pricing model, key features, target audience, biggest strength, and main weakness. Present the results in a comparison table, followed by a 3-bullet summary of where the market gaps are.

Replace [your industry here] with your actual industry. For example: "project management SaaS," "AI writing tools," "e-commerce platforms."

Step 3: Watch the Agents Work

Click Create Task. The Researcher agent starts gathering data immediately. You'll see its output stream in real time.

When the Researcher finishes, the Writer agent takes the raw findings and produces a formatted report. The whole process takes 30–90 seconds.

Total time from signup to your first result: under 3 minutes.

What the Output Looks Like

Here's what you'll receive:

Competitor	Pricing	Key Features	Target	Strength	Weakness
Competitor A	$29-99/mo	Dashboard, Reports, API	Mid-market teams	Strong brand	Limited customization
Competitor B	Free / $49/mo	Automation, Integrations	Small business	Easy setup	No advanced analytics
Competitor C	Usage-based	ML Pipeline, SDK	Developers	Technical depth	Steep learning curve

Market gaps:

• No solution targets non-technical teams needing advanced analytics
• Pricing models are either too simple (flat rate) or too complex (usage-based)
• Integration ecosystems are incomplete across all three

This is a real deliverable you can share with your team, include in a strategy doc, or use to inform product decisions.

Cost of this task: approximately $0.03–$0.08.

3 More Tasks to Try Right Now

Once you've completed your first task, try these. Each takes under a minute to set up:

Task 2: Blog Post Outline

Prompt:

Research the top 5 most-searched questions about [your topic]. For each, provide a 2-sentence answer and suggest a blog post angle. Format as a content brief with target keywords.

Expected output: A content brief with 5 blog post ideas, each with target keywords and angles.

Cost: $0.02–$0.05

Task 3: Customer Research Summary

Prompt:

Analyze the common complaints and praise for [your product or a competitor's product] based on publicly available reviews. Group the feedback into categories (pricing, features, support, UX). For each category, summarize the sentiment and quote 2-3 representative comments.

Expected output: A structured sentiment analysis with categorized feedback.

Cost: $0.04–$0.10

Task 4: Email Draft

Prompt:

Write a cold outreach email to a VP of Engineering at a mid-size SaaS company. The goal is to introduce [your product] and get a 15-minute call. The email should be under 150 words, reference a specific pain point (managing multiple AI tools), and include a clear CTA. Write 3 variations with different tones: professional, casual, and bold.

Expected output: 3 email variations ready to send.

Cost: $0.01–$0.03

Why Multi-Agent > Single Chatbot

You might wonder: "Can't I just paste these prompts into ChatGPT?"

You can. But multi-agent squads produce better results for three reasons:

1. Specialized prompts. Each agent has a focused system prompt optimized for its role. The Researcher is trained to be thorough. The Writer is trained to be clear. A single chatbot has to be mediocre at both.

2. Structured handoffs. The Researcher produces raw data. The Writer transforms it into polished output. This pipeline produces cleaner results than asking one AI to do both steps in the same conversation.

3. No context decay. Long single-chatbot sessions lose focus. Each agent in a squad starts fresh with a clear objective. Quality stays high no matter how many tasks you run.

How to Scale Beyond Your First Task

After completing 3–4 tasks, you'll start seeing patterns in what you need. Here's how to level up:

Add a Reviewer Agent

Add a Reviewer role to your squad. This agent checks the output of your Researcher and Writer for accuracy, completeness, and formatting. It's a quality gate that catches errors before you see them.

Build Specialized Squads

Create separate squads for separate workflows:

• Research Squad: Researcher + Data Analyst + Writer (for deep research)
• Content Squad: Researcher + Writer + Reviewer (for publishing-quality content)
• Sales Squad: Researcher + Writer (for outreach and account research)

Connect Your Local Agents

If you use Claude Code or OpenCode, connect them directly to Ivern with one command:

npx @ivern-ai/agent install --key ivern_YOUR_KEY --provider claude

Your local agent becomes a squad member. It picks up tasks from Ivern and submits results automatically. See the Getting Started guide for the full setup.

Frequently Asked Questions

How much does this actually cost?

The Ivern platform is free. You pay only for AI API usage at direct provider pricing. Most users spend $3–5 per month for regular use. A single task costs $0.01–$0.15 depending on complexity.

Do I need to know how to code?

No. Everything in this tutorial works in your browser. The squad builder, task assignment, and output review are all web-based. If you want to connect local CLI agents like Claude Code, there's an install command — but it's optional.

What if the output isn't good enough?

Refine your prompt. Be more specific about what you want, the format, the length, and the audience. The more context you give, the better the output. You can also iterate — give the agent feedback and ask it to revise.

Can I use this for my specific industry?

Yes. The agents are general-purpose — they adapt to whatever context you provide in your task prompt. Replace [your industry] with anything: healthcare, fintech, real estate, education, e-commerce, etc.

What's the difference between BYOK and BYOA?

BYOK (Bring Your Own Key): You add your Anthropic or OpenAI API key. Ivern calls the models directly. Zero markup.

BYOA (Bring Your Own Agent): You connect a local agent like Claude Code or OpenCode. It runs on your machine and polls Ivern for tasks. Ivern just coordinates.

Both paths land in the same squad task board. You can mix them in the same squad.

Start Now

You've read the guide. Now do the thing.

1. Get your API key (90 seconds)
2. Sign up at ivern.ai/signup (30 seconds)
3. Run the competitor analysis prompt from this guide (60 seconds)

Your first result is 3 minutes away.

For more workflows, check out 10 AI Agent Workflows You Can Set Up Today. For the technical setup, read the Getting Started tutorial.

Create Your Free Squad →