Korea Government Grants Scraper — Bizinfo (기업마당)

What does this Actor do?

This Actor scrapes currently-open Korean government support programs from Bizinfo (bizinfo.go.kr), the official SME support portal operated under Korea's Ministry of SMEs and Startups. Every year thousands of grants, subsidies, export programs, and R&D supports for Korean businesses are announced there — and the portal has no public API.

For each announcement you get the title, application period, parsed deadline with days-left, the responsible agency, and attachment download links. Turn on one switch and the Actor also extracts the full text of HWPX and PDF attachments — which matters, because the actual eligibility criteria (region, business type, revenue caps, exclusions) are almost never on the listing page. They live inside the attachments.

Why use it?

• No API exists. Bizinfo is listing-page-only; even its rows parameter is ignored (15 items per page, always). This Actor handles the pagination for you.
• Deadlines as data. 2026.07.28 ~ 2026.08.14 becomes deadline: "2026-08-14", daysLeft: 11. Always-open programs are labeled explicitly.
• The attachments are the point. In a 57-announcement field test, 56% of announcements had disqualifying eligibility rules that appeared only in the attachments — invisible on the listing. This Actor is the only scraper we know of that reads them (HWPX via built-in parser, PDF via poppler).
• Honest failure reporting. Legacy binary .hwp (HWP 5.x) and scanned-image PDFs cannot be extracted — each one is reported by name with the reason, never silently skipped.

How to use it

1. Set Max announcements (default 30, newest first; up to ~1,500 open programs exist at any time).
2. Optionally enable Extract attachment text — slower, but this is where the eligibility rules are.
3. Run, then download the dataset as JSON, CSV, or Excel.

Input

{
    "maxItems": 30,
    "includeAttachmentText": true
}

Output

{
    "pid": "PBLN_000000000125031",
    "title": "2026년 빅웨이브 글로벌(일본 5차) 참가기업 모집 공고",
    "url": "https://www.bizinfo.go.kr/sii/siia/selectSIIA200Detail.do?pblancId=PBLN_000000000125031",
    "applicationPeriod": "2026.08.04 ~ 2026.08.21",
    "deadline": "2026-08-21",
    "daysLeft": 15,
    "agency": "중소벤처기업부",
    "attachments": [{ "atchFileId": "FILE_000000000767…", "fileSn": "0", "url": "…" }],
    "attachmentText": "…full text of the announcement attachments…",
    "attachmentTextLength": 9602,
    "attachmentErrors": [{ "file": "FILE_…", "error": "legacy binary HWP 5.x — not extractable" }]
}

field	meaning
deadline / daysLeft	Parsed end date; null for always-open programs (see applicationPeriod)
attachmentText	Concatenated text of all readable attachments (HWPX + PDF, zip archives searched one level deep)
attachmentErrors	Per-file failures with reasons — HWP 5.x binaries and scanned PDFs are known limits

Pricing

Charged per announcement scraped. Attachment extraction adds platform usage (downloads) but no extra per-event fee.

Limitations & fair use

• Requests are rate-limited (0.5 s apart) out of courtesy to the public portal. A 30-item run with attachments takes a few minutes.
• Legacy binary .hwp (pre-2014 format) and scanned-image PDFs are reported, not extracted.
• The data is public-sector information published for businesses; re-check the original announcement before applying — programs get amended.

Related

Pairs with HWPX Text Extractor (same developer) if you only need the document-parsing half.