I Built a Customer Tracker Without Hiring a Developer

Every project gets a real database. Your agent writes the SQL. You get the data.

You need to track something. Customer signups. Feature requests. Bug reports. Revenue by plan.

The old way: hire a developer. Set up a database. Write migrations. Build an admin panel. Pay $5,000. Wait three weeks.

The new way: tell your agent "I need a table to track customer signups." It creates it. You query it. Done.

The Problem

Every founder hits this wall. You have data. You need to store it. You need to query it. You need to make decisions from it.

But databases are scary. There's Postgres, MySQL, MongoDB. There's migrations, backups, indexing. There's "have you considered sharding?" from that one friend who read a blog post about it.

You're not a DBA. You're a founder. You just want to know how many users signed up this week.

The Solution

Penut gives every project its own Postgres database. Your agent writes the SQL. You approve the changes. The data is yours.

Need a new table? Your agent creates it. Need to add a column? Your agent adds it. Need to run a query? Your agent runs it.

No DevOps. No database admin. No "let me check with the engineering team."

What You Get

A real database, zero setup. Every project gets an isolated Postgres schema. Not a spreadsheet. Not a Notion database. A real, queryable, indexable database.

Your agent writes SQL. You describe what you need in plain English. Your agent writes the query. You review it. It runs.

Safe by default. Destructive operations — dropping tables, deleting without filters — require your approval. Your agent can't accidentally nuke your data.

Export anytime. Need the data in a CSV? Your agent exports it. Need JSON? Done. No vendor lock-in.

Schema introspection. Your agent can describe your tables, find dependencies, and suggest improvements. It knows your data better than you do.

Real-World Example

You're launching a waitlist. You need to track emails, signup dates, and referral sources.

You tell your agent: "Create a waitlist table with email, signup date, and referral source." It writes the SQL. You approve. The table exists.

Two days later: "How many signups came from Twitter this week?" Your agent writes the query. You get the number.

Next week: "Export the waitlist as CSV so I can import it into my email tool." Done.

Total developer cost: zero. Total time: 10 minutes.

What This Means for You

Stop paying for databases you don't use. You already have one. It's built in.

Ask questions, get answers. "How many users churned this month?" Your agent knows. No dashboard setup required.

Move fast without breaking things. Approval gates protect your data. Your agent can suggest changes, but you make the call.

Own your data. It's in Postgres. It's yours. Export it anytime. No proprietary format. No "your data is trapped in our system."

Next Step

Tell your agent to create a table for something you've been tracking in a spreadsheet. Watch it happen in seconds.

Ready to fire your DBA? Kidding again. But seriously — [sign up for Penut](https://penut.ai) and give your project a database today.

Next up: My Agent Auto-Replies to Emails While I Sleep — how Penut's event system handles incoming messages automatically.