NIH Type1 Territory Award Review

Regional scientific supplier sales operations teams can retrieve a small factual queue of NIH Type1 administrative award notices for human CRM account review. Make one weekly API call for your owned states and administering NIH institutes, then supply previously reviewed applId values to omit those exact records on later runs.

The Actor uses the official, keyless NIH RePORTER Projects API. It checks the actual returned award type, project-number application type, organization state, administering institute and notice calendar date before returning each award. Type1 is a source administrative application classification; it does not mean the institution is a new NIH customer.

Try it

The form prefill, sample-input.json and API exampleRunInput use the same input:

{"states":["MA"],"adminInstitutes":["NCI"],"maxResults":20,"previousApplicationIds":[]}

At runtime the omitted dates become UTC today minus 30 days through UTC today, inclusive. The Actor filters the source notice calendar date while preserving its original timestamp. It searches all fiscal years, including when the interval crosses October 1, and does not enable an active-only filter.

Optional awardFrom and awardTo are valid YYYY-MM-DD dates, inclusive, with a maximum 30-day span and no future awardTo. If supplying one date, the other still takes its ordinary runtime default; supply both for a historical window. There are no fixed dates in prefill or the API example.

Input limits: 1–5 US states/DC, 1–3 supported administering institutes, 1–100 maximum new output awards, and at most 1000 positive integer previous application IDs. State and institute abbreviations are uppercase. Territories are outside this MVP. Supported NIH extramural ICs: NCI, NEI, NHLBI, NHGRI, NIA, NIAAA, NIAID, NIAMS, NIBIB, NICHD, NIDCD, NIDCR, NIDDK, NIDA, NIEHS, NIGMS, NIMH, NIMHD, NINDS, NINR, NLM, NCCIH, NCATS, FIC. FDA/AHRQ, OD, intramural-only centers and historical NCRR are outside this product's allowlist.

Results and coverage

Each award row contains stable applId, project number, both source application-type fields, organization name/state, UEI/IPF if present, administering institute, source award amount, raw notice and project-start dates, source fiscal year, NIH project URL, observation time and newToSuppliedIds: true. Missing optional values are null. awardAmountUSD is the administrative source award amount, not a future grant total, available purchasing budget or spending forecast.

Identical application IDs are deduplicated. Conflicting records for the same application ID are omitted. Already supplied application IDs incur no custom result fee. No record's absence implies termination, expiry, withdrawal or no change.

Retrieval is capped at two sequential pages of 100 source records, 25 seconds and 2 MB per response, with at least 1.1 seconds between request starts. Output is further bounded by maxResults and the run's available award budget. Every row includes coverage: returned source total, retrieved pages/rows, invalid/out-of-scope and duplicate counts, prior-ID omissions, output count and truncation reason. A retrieved-scope count is never a claim of complete NIH coverage. Source updates, lag and changes while paging remain possible. All fiscal years are requested; current FY2027 availability is not guaranteed by this filter.

No useful rows produces one summary row with zero custom award fee. On missing/error/403/429/redirect challenge data, a source_failed_or_unknown summary is saved and the run fails with no custom award fee. There are no source retries or challenge bypasses. A later page error releases no first-page award rows.

The default Dataset is the saved review queue. Key-value OUTPUT contains authoritative coverage and custom charge settlement. CHECKPOINT records the same funded slice and operation state. A native restart reads the checkpoint, returns a fully settled result without republishing, and refuses automatic replay of an incomplete publication or unknown remote charge. A failed run with an ambiguous charge requires native ledger reconciliation by the release/maintenance operator; it never blindly charges again. Dataset rows preserve the publication-time billing state; use OUTPUT and the platform charge ledger for settlement.

Price

Proposed Pay-per-event pricing: $0.005 per unique eligible award actually saved, plus a separate automatic $0.00005 Actor start event. Default Dataset item fee is zero. Price configuration must match exactly at runtime; an unexpected event fee is rejected before source retrieval. At most 20 default awards means up to $0.10005 in event charges; the 100-row cap means up to $0.50005. Normal Apify platform terms, compute treatment and any account/owner fee waiver remain platform-controlled. A local unbilled test does not prove native owner-waiver behavior. Native pricing and actual compute are unverified before the first authorized cloud run.

The funded award capacity is applied before any useful Dataset, OUTPUT or CHECKPOINT save. The funded slice is persisted and published before per-award charging. Zero, subunit and fractional limits exclude unfunded rows. Source errors never create custom award events. A publication/charge interruption can leave accessible funded rows whose settlement is unknown or incomplete; the checkpoint marks that state and prevents replay. Automatic startup charges are separate: zero custom awards does not mean a totally free native run.

Source and use

Source attribution: NIH RePORTER; observation time appears on each award. The official API is intended for third-party reporting/integration/business applications and documents its access limits. This Actor uses only administrative facts and NIH record links, under the selected-source/general public-domain basis recorded for this project. Government general guidance is not a blanket license for third-party works. Titles, abstracts, PI names, contacts, scientific interpretation, equipment-need guesses, clinical decisions and third-party images/logos are excluded. No NIH endorsement is implied. Human account review is required.

sample-output.json is a dated actual local observation, not a perpetual freshness guarantee. The official NIH Awards by Location & Organization also explains that current data can change. This product's results are an account-review input and do not establish any organization's purchasing intent.

Development

Requires Node 24.18.0 and the pinned Apify SDK 3.7.2. npm ci --ignore-scripts, npm test, npm run validate, npm run live. live makes a bounded real official API request, writes a minimized administrative fixture/sample and verifies actual local SDK INPUT/Dataset/OUTPUT/CHECKPOINT storage. It performs no native build/run/publish or paid upstream call. Tests clear cloud credentials/flags and set a fresh SKU-local resolved storage path before the first SDK import, assert the actual SDK storage location and never purge/drop shared storage. Production local launches likewise use fresh SKU-local storage; native restarts use native run storage.

The Docker image installs from package-lock.json. Local dependency copies, npm cache, storage and tests are excluded from the Docker context; a successful native Docker build and schema recognition still need independent release verification. This MVP contains native fetch and Apify storage/charging only; it has no notifications, contact enrichment, browser, proxy, LLM, database or dashboard.