NSF Awards Scraper - Research Grants, PIs & Funding Data

Extract NSF award records from the official research.gov Awards API. Returns rich structured data including award title, abstract, awardee organization, principal investigator (name + email), NSF program officer (name/email/phone), funds obligated and estimated totals, directorate and division, program element/reference codes, CFDA numbers, start/expiration dates, and performance location. Filter by keyword, PI name, awardee organization, state, date range, and CFDA program.

What You Get

Each record contains:

• Award identification: award ID, title, CFDA number, directorate/division abbreviations, program element and reference codes
• Awardee details: organization name, city, state, zip, country, UEI number
• Principal Investigator (PI): first/last/full name, email address
• Co-PIs: semicolon-separated list of co-PI names
• Program Officer (PO): name, email, phone — the key contact at NSF
• Funding: funds obligated (USD), estimated total (USD)
• Dates: start date, expiration date, latest amendment date
• Program context: NSF funding program name, primary program, org long name
• Performance location: city, state, zip, country (may differ from awardee)
• Public access mandate flag
• Direct URL to the NSF award detail page

Use Cases

Buyer	What They Use It For
University sponsored-programs offices	Track competitor institutions' NSF funding; identify gaps in their own programs
Grant writers	Research NSF program officers by directorate; identify active programs and funded PIs
Biotech / deep-tech VCs	Identify university researchers with NSF backing as potential spinout founders
Research-policy analysts	Analyze funding flows by directorate, state, or institution over time
Competitive-intel teams	Track who NSF is funding in a specific technology area
Lead-gen / BD teams	PI and PO email fields for direct outreach in the academic/government sector

Input Parameters

Parameter	Type	Description
keyword	string	Free-text search across award title and abstract
piName	string	Filter by PI name (e.g. Smith)
awardeeName	string	Filter by awardee institution (e.g. Massachusetts Institute of Technology)
awardeeStateCode	string	Filter by US state code (e.g. CA, MA, NY)
dateStart	string	Earliest award start date, MM/DD/YYYY format
dateEnd	string	Latest award start date, MM/DD/YYYY format
cfdaNumber	string	CFDA program number (e.g. 47.070 for Computer & Information Science)
maxItems	integer	Maximum records to return. Default 10. Set 0 for unlimited (use date slicing for large pulls).

Handling the API's 3,000-Record Limit

The NSF Awards API returns at most ~3,000 results per query. For larger data pulls:

1. Slice by year: run one query per calendar year using dateStart/dateEnd
2. Combine with keyword or awardeeStateCode to narrow each slice further

Example: to get all 2024 awards in California, set dateStart: 01/01/2024, dateEnd: 12/31/2024, awardeeStateCode: CA.

Sample Output Record

{
  "award_id": "2606034",
  "title": "CAREER: Generalization Capabilities of Machine Learning...",
  "awardee_name": "University of California-Los Angeles",
  "awardee_city": "LOS ANGELES",
  "awardee_state_code": "CA",
  "uei_number": "RN64EPNH8JC6",
  "pi_full_name": "Hayden Schaeffer",
  "pi_email": "hayden@math.ucla.edu",
  "po_name": "Stacey Levine",
  "po_email": "slevine@nsf.gov",
  "po_phone": "7032922948",
  "funds_obligated_amt": 214876,
  "estimated_total_amt": 214876,
  "start_date": "10/01/2026",
  "exp_date": "09/30/2029",
  "cfda_number": "47.049",
  "directorate_abbr": "MPS",
  "division_abbr": "DMS",
  "fund_program_name": "APPLIED MATHEMATICS",
  "program_element_codes": "126600",
  "program_reference_codes": "075Z, 079Z",
  "public_access_mandate": "1",
  "perf_state_code": "CA",
  "award_detail_url": "https://www.nsf.gov/awardsearch/showAward?AWD_ID=2606034",
  "status": "success"
}

Data Source

All data comes from the NSF Awards API (research.gov), the official public interface to the NSF Awards database. No authentication required. The API covers 500,000+ awards since 1959.

The scraper requests all 40+ available fields by name, including fields like piEmail, poEmail, abstractText, and coPDPI that the API only returns when explicitly listed in printFields — ensuring no data is silently omitted.

Technical Notes

• Rate limiting: the scraper sends ~1 request per second as per API courtesy guidelines, with automatic retry on 429/503 responses
• Memory: 512 MB default, sufficient for full paginated runs
• Timeout: 4-hour ceiling for full-history slices; typical keyword searches complete in seconds
• No proxy required: the NSF Awards API is public and accessible without geo-routing