> ## Documentation Index
> Fetch the complete documentation index at: https://docs.datacircle.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Installation

> Download and set up the DataCircle leads dataset on your machine.

# Installation

## 1. Create an account

Go to [datacircle.dev](https://datacircle.dev) and sign up with your email.

## 2. Download the data

Once signed in, click **Download** to get the dataset. The download is a zip archive containing CSV files for companies and people.

## 3. Unzip the file

```bash theme={null}
unzip datacircle-leads.zip -d datacircle-leads
```

This creates a folder with:

* `companies.csv` - 1,752,762 company records
* `people.csv` - 11,139,026 person records

## 4. Load into a database

Point your AI agent (ChatGPT, Claude, Cursor, etc.) to this documentation page and ask it to set up the database for you. It will handle the rest.

If you prefer to do it manually, here are two options:

### Option A: DuckDB

[DuckDB](https://duckdb.org) is a lightweight analytical database that runs locally with zero setup.

```bash theme={null}
pip install duckdb
```

```python theme={null}
import duckdb

con = duckdb.connect("datacircle.duckdb")

con.execute("""
    CREATE TABLE companies AS
    SELECT * FROM read_csv_auto('datacircle-leads/companies.csv')
""")

con.execute("""
    CREATE TABLE people AS
    SELECT * FROM read_csv_auto('datacircle-leads/people.csv')
""")

# Find all software companies in California
result = con.execute("""
    SELECT NAME, URL, EMPLOYEE_COUNT_RANGE
    FROM companies
    WHERE LINKEDIN_INDUSTRY = 'Computer Software' AND HQ_STATE_CODE = 'CA'
    ORDER BY LINKEDIN_FOLLOWERS DESC
    LIMIT 10
""").fetchdf()

print(result)
```

### Option B: ClickHouse

[ClickHouse](https://clickhouse.com) is a high-performance analytical database, ideal for larger datasets.

```bash theme={null}
curl https://clickhouse.com/ | sh
./clickhouse server
```

```sql theme={null}
CREATE TABLE companies
ENGINE = MergeTree()
ORDER BY NAME
AS SELECT * FROM file('datacircle-leads/companies.csv');

CREATE TABLE people
ENGINE = MergeTree()
ORDER BY LAST_NAME
AS SELECT * FROM file('datacircle-leads/people.csv');

-- Find VPs at mid-market companies
SELECT p.FIRST_NAME, p.LAST_NAME, p.CURRENT_JOB_TITLE, c.NAME AS company
FROM people p
JOIN companies c ON p.CURRENT_JOB_COMPANY_LINKEDIN_ID = c.LINKEDIN_ID
WHERE p.CURRENT_JOB_TITLE LIKE '%VP%'
  AND c.EMPLOYEE_COUNT_RANGE IN ('51-200', '201-500')
LIMIT 10;
```

## 5. Start exploring

You now have 11M+ U.S. B2B leads at your fingertips.

See the [Field Reference](/data/fields) for the complete schema and fill rates.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.