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

# Field Formats

> Format specifications for Data Legion data fields: cleaned vs raw text variants, date formats, phone number formatting, ISO codes, and more.

Detailed format specifications for all field types.

## Cleaned vs Raw Variants

All text data points include both cleaned (normalized) and raw (source) variants:

* **Cleaned**: Standardized, normalized text for consistent filtering. Lowercased, accents stripped, symbols normalized.
* **Raw**: Original source formatting preserved. Contains all source variations as an array.

**Person example** (job title):

```json theme={null}
{
  "title": {
    "cleaned": "senior product manager",
    "raw": [
      "Senior Product Manager",
      "Sr. PM",
      "Sr Product Manager"
    ]
  }
}
```

**Company example** (company name):

```json theme={null}
{
  "name": {
    "cleaned": "stripe inc",
    "display": "Stripe",
    "raw": [
      "Stripe, Inc.",
      "Stripe"
    ]
  }
}
```

Use `cleaned` for filtering and matching. Use `raw[]` to see original source data. Company `name` additionally includes `display` (premium tier) — the best-quality name with original casing preserved, ideal for UI rendering.

In person data, cleaned/raw is used for: job titles, organization names, headlines, summaries, skills, languages, degrees, fields of study, certification names, and certification institutions. In company data, it is used for: company name (plus a `display` variant), headline, and description.

## String Fields

All string fields are:

* Lowercased (except where case matters)
* Trimmed of leading/trailing whitespace
* Normalized for accents and special characters

## Date Fields

Dates preserve precision and use ISO 8601 format:

| Format       | Example      | Description    |
| ------------ | ------------ | -------------- |
| `YYYY`       | `2020`       | Year only      |
| `YYYY-MM`    | `2020-03`    | Year and month |
| `YYYY-MM-DD` | `2020-03-15` | Full date      |

When multiple sources provide different precisions for the same date, we keep the most specific version.

## Phone Numbers

All phone numbers are in E.164 international format:

* Format: `+[country code][number]`
* Example: `+15551234567` (US number)
* Example: `+442071234567` (UK number)

All phone numbers include the country code prefix.

## Email Addresses

All email addresses are:

* Lowercased
* Trimmed of whitespace
* Checked for syntax/format

Example: `JANE.DOE@COMPANY.COM` → `jane.doe@company.com`

## Geographic Codes

### State Codes

ISO 3166-2 format:

* `US-CA` (California, United States)
* `US-NY` (New York, United States)
* `GB-ENG` (England, United Kingdom)

### Country Codes

ISO 3166-1 alpha-2 format:

* `US` (United States)
* `GB` (United Kingdom)
* `CA` (Canada)

### Continent Codes

Standard continent codes:

* `NA` (North America)
* `EU` (Europe)
* `AS` (Asia)
* `SA` (South America)
* `AF` (Africa)
* `OC` (Oceania)
* `AN` (Antarctica)

## URLs

All URLs are normalized:

* Protocol included (`https://`)
* Trailing slashes removed (except root paths)
* Consistent formatting

Examples:

* `linkedin.com/in/janedoe` → `https://www.linkedin.com/in/janedoe`
* `github.com/user` → `https://github.com/user`

## Boolean Fields

Boolean fields use `true`/`false` values:

* `is_decision_maker`: `true` or `false`
* `current`: `true`, `false`, or `null` (for personal emails)
* `validated`: `true` or `false`

## Integer Fields

Integer fields are numeric:

* `age`: Years as integer
* `birth_year`: Year as integer
* `years_of_experience`: Years as integer

## Enum Fields

Enum fields use lowercase, underscore-separated values:

* `seniority_level`: `c_level`, `vp`, `director`, etc.
* `job_function`: `engineering`, `sales`, `marketing`, etc.
* `confidence`: `high`, `moderate`, `low`

See [Enum Values](/docs/enums) for complete lists.

## Array Fields

Arrays contain objects with consistent structure.

**Person arrays:**

* `phones[]`: Array of phone objects
* `emails[]`: Array of email objects
* `locations[]`: Array of location objects
* `experience[]`: Array of experience objects
* `education[]`: Array of education objects
* `socials[]`: Array of social profile objects
* `skills[]`: Array of skill objects
* `languages[]`: Array of language objects

**Company arrays:**

* `domains[]`: Array of domain objects
* `socials[]`: Array of social profile objects
* `tickers[]`: Array of ticker objects
* `legion_employee_count_by_month[]`: Array of monthly headcount snapshots

All arrays are sorted deterministically. See [Array Ordering](/docs/data-standardization/array-ordering) for details.

## Related Documentation

* **[Array Ordering](/docs/data-standardization/array-ordering)** - How arrays are sorted
* **[Enum Values](/docs/enums)** - All possible enum values