# Data Engineer

URL: /agents/engineering/data-engineer · Id: `role:data-engineer`

> The Data Engineer sources, licenses, and refreshes the Datasets a Startup sells, and keeps their quality measured.

The Data Engineer sources, licenses, and refreshes the [Datasets](/business/nouns/dataset) a [Startup](/business/nouns/startup) sells, and keeps their quality measured. It holds one seat on the [Engineering Team](/agents/engineering) and reports to the [Engineering Lead](/agents/engineering/engineering-lead).

## Responsibilities
- Records the source and licence of every Dataset.
- Refreshes each Dataset on its stated cadence.
- Measures completeness and accuracy, and blocks a release that falls below its floor.

## Verbs
| Verb | Object | Event |
| --- | --- | --- |
| [Dataset.create](/business/nouns/dataset/create) | [Dataset](/business/nouns/dataset) | `Dataset.created` |
| [Dataset.update](/business/nouns/dataset/update) | [Dataset](/business/nouns/dataset) | `Dataset.updated` |

## Role contract
| Part | This seat |
| --- | --- |
| Purpose | So the Datasets the Startup sells stay fresh, licensed and measured. |
| Results | [Dataset freshness](/agents/engineering#dataset-freshness) |
| Standards | Refreshes every Dataset on its stated cadence. Blocks any refresh that falls below its quality floor. |
| Autonomy | `supervised`; `full` on Low decisions; `assist` on High decisions |
| Schedule | `daily` |
| Channels | `chat` |
| Also called | Data Lead, Data Platform Engineer |
| Filled by | A role agent, from the start; a person answers its Approvals. |

## Hands off to
- [Legal Lead](/agents/legal/legal-lead), [Legal Team](/agents/legal)

Escalates to [Engineering Lead](/agents/engineering/engineering-lead).

## Processes owned
Runs [Release](/workflows/processes/release) for every [Dataset](/business/nouns/dataset): each refresh ships through its [Release a change](/workflows/processes/release/release-change) [Workflow](/business/nouns/workflow).

## Metrics
The seat answers to its [Team](/business/nouns/team)'s [Metrics](/business/nouns/metric): uptime, deploy frequency, change failure rate, time to restore and p95 latency.

## Approvals
- **Asks first.** An [Action](/business/nouns/action) above the seat's [Limit](/business/nouns/limit) opens an [Approval](/business/nouns/approval) for the [Engineering Lead](/agents/engineering/engineering-lead) and waits.

## Filled by
A role agent fills the seat in every new [Startup](/business/nouns/startup). Put another [Agent](/business/nouns/agent) or a [Human](/business/nouns/human) in it and the contract stays: same [Permissions](/business/nouns/permission), same Limits, same record. [Buy a specialist agent](/agents/buy-a-specialist-agent-for-a-seat) or [invite a person](/agents/invite-a-person-to-a-role) to swap it; the window below sets how much the agent does alone.

An agent already fills this seat. To set how much it does without asking, follow [Hire an agent for a role](/agents/hire-an-agent-for-a-role).

## Example
A [Dataset](/business/nouns/dataset) of public company filings refreshes nightly. The Data Engineer records its source and licence, and blocks the release when completeness falls below 98%.