Delivery options
Choose the destination that fits your research workflow; delivery is configured with our team.
| Destination | Access setup | Delivery |
|---|---|---|
| S3 | Agreed bucket or delivery location and scoped IAM access. | Recurring files with agreed format, history, cadence, and snapshot retention. |
| Snowflake | Agreed account, database/schema, and scoped service and consumer roles. | A managed delivery pipeline into warehouse tables; provisioning and refresh are coordinated during onboarding. |
Contact us for S3 delivery · Contact us for Snowflake delivery.
Snowflake is a delivery pipeline for now. File layout, destination setup, service levels, and supported refresh cadence are confirmed for each contract.
Datasets and warehouse tables
Use the dataset references for definitions, fields, identifiers, and interpretation.
- Operating KPIs: company-reported metrics, guidance, definitions, and source references.
- Job Postings: hiring observations, dimensions, classifier versions, and continuity metadata.
- Adoption Signals: mapped projects, artifacts, versions, events, and adoption activity.
Adoption Signals warehouse design describes the four activity tables and proposed metadata dictionaries, enriched views, and publication manifest. These additions are a delivery design, with rollout coordinated separately.
Set up delivery
- Contact us with the datasets, company coverage, intended use, destination, history, and refresh cadence you need.
- Agree on dataset licensing, permitted users and uses, delivery scope, and support terms.
- Configure scoped S3 access or the Snowflake destination and service role. Exchange credentials through the agreed secure setup process.
- Validate a representative delivery against the dataset schema, identifiers, coverage, null semantics, and publication metadata.
- Enable the recurring pipeline and agree on monitoring, correction handling, schema changes, and snapshot retention.
Do not send credentials through the contact form. The initial inquiry only needs your requirements and destination preference.
Refresh and research conventions
- Freshness and history depend on the dataset and source coverage. Use the delivered publication and coverage metadata rather than assuming every series updates together.
- Preserve nulls and availability fields. Missing observations are not zero activity.
- Keep identifiers, classifier/schema versions, and source references with extracts so results can be reproduced.
- Maintained history can receive corrections. Agree on dated snapshots and retention if your research requires point-in-time reproducibility.
- Validate complete publications before replacing a prior delivery. Publication consistency and retry behavior are part of destination onboarding.
For agent access instead, see MCP setup and contact us from the MCP page in Research.