This lead-gen AI agent works while you sleep — here’s the setup

Outbound is a pain when you’re doing everything yourself.

Here’s how to build an AI agent that does it for you — one that quietly finds high-quality leads based on real buying signals and sends them to you every day. You’ll use a few tools (no code), and you’ll be able to set the whole thing up in an hour or two.

Let’s get into it.

What you’re going to build

A system that:

• Finds companies that match your ideal customer profile
• Checks if they use certain tools (like Apache Airflow, for example)
• Looks for real buying signals — new hires, pricing pages, tech stacks
• Finds decision-makers at those companies
• Sends everything to a spreadsheet or Airtable
• (Optionally) enriches and emails them

And it runs without you lifting a finger once it’s set up.

Step 1: Get crystal clear on who you’re trying to reach

Start with this.

Before you touch any tools, write down what your ideal customer looks like.

Think:

• Company type
• Pricing model
• Tech stack
• Team size
• Job titles that make decisions
• Signals that show buying intent (e.g. new funding, hiring)

For example: “I want to find SaaS companies with public pricing under $50/month that use Apache Airflow and have recently hired a Head of Engineering.”

This sentence becomes your filter logic. It’s what you’ll feed into the agent later.

Step 2: Pick your tools

Here’s what we’ll use:

• Origami Agents — AI-powered browser agent builder
• Airtable — lead database
• Clay.run — for enrichment + email outreach (optional)
• Google Sheets — simpler Airtable alternative if needed

All of these are low-code. If you can use Notion, you can use these.

Note: Origami integrates directly with both Airtable and Google Sheets — no manual export/import needed.

Step 3: Create your agent in Origami

Head to Origami, sign up, and click “Create Your Agent.”

Give it a name like: \\\ Airflow SaaS Lead Finder \\\

You’ll now see a visual builder — kind of like Zapier, but with a browser and brain behind it; each step does something specific.

Define your filters

You want the agent to look for companies that meet specific criteria. In Origami, stack filters like:

• “Uses Apache Airflow”
• “Pricing page under $50/month”
• “Recently hired Head of Engineering”

You can combine filters — the more specific, the better.

Origami will crawl company sites, LinkedIn, Crunchbase, and blogs to look for these patterns.

Add what to extract

Now tell Origami what to pull once it finds a match:

• Company name
• Website
• Tech stack mentions
• Pricing info
• Decision-maker name + title
• LinkedIn URL
• Contact email (if available)
• Signal it matched on

This becomes your lead record.

Step 4: Send the data somewhere useful

Origami will ask where to push the results.

Choose Airtable or Google Sheets. Airtable gives you more flexibility long-term (grouping, filtering, tagging). But both work.

Origami will walk you through connecting your account and mapping each field visually.

Now, when your agent finds a match, it’ll send the lead straight to your database.

Step 5: Let it run

Click “Run Agent.”

You’ll see Origami open browser tabs, read pages, extract data, and send matches to your sheet. The first run might take 10–15 minutes depending on the filters.

You can schedule it to run daily or weekly. I recommend starting with once per day.

Now you’ve got a background agent quietly scouting for new leads while you focus on your product.

Step 6: Review and tune

Once the first batch of leads comes in, look closely.

Ask:

• Are these the kind of companies I want?
• Are the decision-makers relevant?
• Are the signals accurate (tech stack, pricing)?

If not, tweak your filters and rerun. You’ll usually get solid output within 1–2 passes.

Remember: You’re not going for hundreds of leads. You’re going for relevance. That’s the whole point of building this in the first place.

Optional: Enrich and email with Clay

Once you’re getting leads you like, you can plug them into Clay.

Clay connects to Airtable and lets you do a few smart things:

1. Enrich the lead — find work email, recent company news, LinkedIn activity
2. Write warm emails based on those enrichments
3. Send 3–5 messages/day, automatically

Here’s an example message you might use: “Hey {{FirstName}}, saw you’re using Airflow at {{Company}}. I’m building something that simplifies \[X\] — mind if I send you a quick link?”

Keep it personal. You’re starting conversations — not campaigns.

That’s it! You just set up a smart lead gen agent that runs quietly in the background, so you can stay focused on what matters.