Building a Startup Launch Team using CompleteAI

Building a Startup Launch Team using CompleteAI

Learn how you can create 3 agents that work together to manage a SaaS product.

We are going to Build 3 collaborating AI Agents that work together to take a startup idea from concept to launch-ready; including market research, product definition, pitch deck, and landing page copy.

What you'll build:

• Research Agent — Analyzes competitors, identifies market gaps, summarizes findings
• Product Agent — Defines features, user flows, and technical requirements
• Marketing Agent — Creates pitch deck content, landing page copy, and go-to-market strategy

Prerequisites:

• Complete.dev account
• Basic familiarity with Agent Builder

Step 1: Set Up Your Shared Workspace:

1. Create a new workspace in Complete.dev called "Startup Launch Team"
2. FIll in the Space Name, Description & members that you would like to add.
3. Invite your teammates (if any)

Step 2: Build the Research Agent:

Go to your dashboard button and click on Add member.

Type in Agent Builder

Add the agent builder and go back to general chat, then send in the prompt:

Create the Agent called "Market Researcher" that is responsible for analyzing competitors, identifying market trends, and finding gaps in the market. This is the system prompt:
You are a market research analyst. Your job is to analyze competitors, identify market trends, and find gaps in the market. Always provide structured output with:

Executive Summary
Competitor Analysis (3-5 key players)
Market Gap/Opportunity
Recommendations
Save all outputs to the /research folder.

But first of all ensure you've inserted your preferred model API key, in our case we will use Gemini API.

Go to google ai studio and get your API Key there, then send it as part of the prompt:

GOOGLE_API_KEY=xxxxxxxx

Once it's done, this is what you'll see as the output.

When you go to Code --> General --> market_researcher.py you can see and verify the agent code that has just been created for your Researcher Agent.

Testing your market research agent:

Agent Builder provides a URL to a live test application, if it hasn't, you can specifically prompt it to do that.

Now you can test the agent using the prompt:

Research the AI scheduling assistant market. Who are the main competitors? What features are missing? Save your analysis to /research/scheduling-market.md

Step 3: Build the Product Agent

Now let's go back to the #general chat and build the product agent. This is the prompt you'll use:

Now let's build the product agent, this is the system prompt: 
You are a product manager and technical architect. Your job is to transform market research into concrete product specifications. Given research input, you will:
- Define core features (MVP scope)
- Create user stories
- Design user flows
- Define technical requirements

Always read relevant research from /research before starting.
Save outputs to /product/ folder.

Once it is done, it will update the same URL. Go to the URL and click on the product agent tab.

Test your agent by sending the prompt

Read /research/scheduling-market.md and create a product spec for an AI scheduling assistant targeting freelancers. Include features, user flows, and tech stack. Save to /product/scheduling-product-spec.md

Step 4: Build the Marketing Agent

Let's then go back to the #general chat, and work on the Growth Strategist agent.

This is the prompt you'll send to the agent builder:

Let's build the Growth Strategist agent, here's the system prompt for it. 

You are a marketing strategist and copywriter. Your job is to create go-to-market materials based on product specs. You will:
- Write compelling landing page copy (headline, benefits, CTA)
- Create pitch deck outline and key slides content
- Define target audience and channels
- Suggest pricing strategy

Read product specs from /product/ before creating materials.
Save outputs to /marketing/ folder.

Once it is done, go to the agent url and refresh the page.

Then head over to the Growth Strategist tab and insert your prompt.

Read /product/scheduling-product-spec.md and create: 
1) Landing page copy for the AI scheduler, 
2) A 5-slide pitch deck outline, 
3) Go-to-market strategy. Save everything to /marketing/ folder.

Making your agents live in the workspace:

Go to your spaces directory.

Then click on the Generate API Key button.

Go back to the #general chat and send the prompt to the agent builder:

Here is my API KEY: sk_xxxxxx, make the agents live in the workspace. 

Once it is done, this is the output you'll see.

We'll then proceed to adding the agents to our workspace officially.

Go to Dashboard --> Add Members.

Then fill in the newly created agents.

Step 5: Orchestrate Multi-Agent Collaboration

Now we're going to demonstrate the agents working together.

You'll proceed to send this prompt to your agents:

@mkt_researcher Research the market for [YOUR IDEA]. Save to /research/market-analysis.md

Once done, @product_agent read /research/market-analysis.md and create product specs. Save to /product/specs.md

Once done, @growth_strat read /product/specs.md and create all marketing materials in /marketing/

And you'll see the final output and how all the agents work with each other.