HELEXPO Exhibitor Companies — Cross-Show Export

Turn HELEXPO's openly published exhibition CSVs into a source-attributed company list. Find suppliers by country or product keyword, exclude domains you already know, and group matching company observations across shows.

Current source: the HELEXPO exhibitions dataset on Growthfund Open Data Hub. This is not a universal trade-show scraper. No login, paid proxy, browser or LLM API is required.

Start with three companies

{
  "mode": "companies",
  "exhibitions": ["AGROTICA 2026"],
  "maxCompanies": 3
}

You get:

• English and original company names, source country, normalized published website and domain.
• Each exhibition occurrence, source-scoped exhibitor ID, product text, pavilion/stand and source row.
• Reported represented entities, clearly distinguished from inferred ownership.
• A spreadsheet CSV, JSON dataset, source attribution and a coverage/error report.

A website marked source_published was listed by the source. It has not been independently visited or verified.

Find available exhibitions first

{"mode": "catalog"}

Catalog mode lists available Exhibitors CSV resources; product-only Exhibits files are excluded. Copy a resource ID or a name fragment including its year into exhibitions. Ambiguous names fail with a useful message rather than silently picking a year.

Examples checked during development include AGROTICA 2026 (agricultural equipment and suppliers), BEYOND 2025 (technology), and FRESKON 2026 (fresh produce). Availability follows the publisher's catalog, not a hard-coded promise of current events.

Input

Field	Default	Meaning
mode	companies	Export companies, or use catalog to discover resources.
exhibitions	["AGROTICA 2026"]	Up to 8 unambiguous exhibition names/fragments or resource IDs.
maxCompanies	100	Output limit after filters and company grouping, 1–10,000.
requireWebsite	false	Only keep companies with a usable source-published website; excludes missing, invalid and social-only URLs. Does not verify the website.
countries	[]	Case-insensitive exact source country names.
keywords	[]	Match any term in company names, product text or represented-entity names.
excludeDomains	[]	Skip known company domains, e.g. example.com.
includeRepresentedCompanies	true	Include source-reported represented entities within occurrences.

Selected CSVs are downloaded in full before the output limit is applied. The source requests a 10-second crawl delay; this is honored. A small run usually takes tens of seconds, not milliseconds.

What grouping means

A company is grouped across resources only when its normalized name, country and registrable website domain agree. Different names sharing a domain remain separate and carry a review note. Rows without a usable company domain remain source-scoped. This intentionally misses some duplicates rather than claiming uncertain legal-entity matches.

Source IDs are scoped to an exhibition resource: exhibitor 001 in two different shows is not, by itself, an identity match. Original names, website strings and source references stay in occurrences.

No estimated revenue, inferred contact emails, personal contacts or automatic outreach is produced.

Output and coverage

On Apify, use the default Dataset and these key-value records:

• COMPANIES.csv — spreadsheet export, with formula-like cell values escaped.
• RUN_SUMMARY — selected resources, source/matched/output counts, errors and coverage.
• ATTRIBUTION.txt — provider, dataset, license and transformation attribution.

Important report fields:

• sourceReadComplete: every selected source was read and required record identities parsed.
• outputCapped: more matching companies existed than maxCompanies allowed.
• complete: no source errors and no output truncation.
• status: succeeded, partial or failed.

Partial source failures preserve usable results but fail the Actor run after saving the report. Invalid websites and malformed optional nested values are explicitly marked in record warnings. An empty filter result is valid; an empty source CSV or a blocked/changed source page is an error. We never invent rows to make a run look successful.

sampledAt is when this Actor downloaded the data. Exhibition year comes from the source record. Neither means the company profile was updated on that date. The publisher's data can be historical.

Pricing

$3 per 1,000 exported companies, plus $0.00005 per run. Pay only for companies delivered in that run and the start event; there is no subscription. The price displayed by Apify at run time is authoritative.

Event	Price
Company exported (company-exported)	$0.003 per company — $3 per 1,000
Actor start (apify-actor-start)	$0.00005 per run at the supported memory limit of 1 GB or less

One company grouped across several exhibitions counts once within a run. New runs may deliver and charge for the same company again. JSON, CSV and source attribution have no separate event fee. There is no additional default Dataset-item fee or platform-usage surcharge.

A company with missing optional fields is still a company result. Use requireWebsite: true to exclude records without a usable published website before billing. This does not buy a verified website, contact person or guaranteed sales lead.

Catalog queries and empty results have no company fee; the startup fee still applies, including to failed runs. Companies successfully delivered before a partial source failure remain billable. The minimum selectable spending budget is $0.01; this is not a minimum charge. The Actor checks the remaining budget and stops delivering companies when it is exhausted.

At these prices, three companies cost $0.00905 and 100 companies cost $0.30005, including one start. Data remains available under its source license: payment is for the hosted processing service, not ownership or exclusive access to the data.

Data license and attribution

Data source: Growthfund_Open Data Hub, Exhibitions - Exhibitors & Exhibits, HELEXPO. The portal terms permit download, reuse and redistribution and specify CC BY 4.0 unless a dataset states otherwise. No override was declared in the checked dataset metadata. Retain the attribution and source/license links when reusing exports. Website normalization, filtering and grouping are our transformations.

This is an independent tool, not endorsed by Growthfund or HELEXPO. Their logos are not used. The data license does not apply to this program's source code.

Troubleshooting and support

• No exhibition matched / ambiguous name: run {"mode":"catalog"} and use the exact resource ID, or include the year in the name.
• No companies match the filters: remove filters or try another exhibition. This can be a successful empty result; it does not imply the source is empty.
• Source HTTP error or changed CSV: check RUN_SUMMARY. Do not assume a partial export is complete or repeatedly retry a blocked source.
• Fewer results than expected: check outputCapped, stoppedByBudget and per-resource counts. Exhibition snapshots may be historical or incomplete at the publisher.
• Greek characters in Excel: use COMPANIES.csv, which includes a UTF-8 BOM. Nested source evidence is preserved most fully in JSON.

When reporting a problem in the Actor's Issues tab, include the run ID, selected exhibition and error message. Do not include account tokens or personal billing information. Start a new run rather than resurrecting a terminated run; resuming a previous delivery is not supported and can duplicate records.

Local development

Python 3.11 or newer and uv are needed locally.

uv sync --frozen
uv run exhibitor-export --input examples/agrotica.json --output-dir outputs/agrotica
uv run exhibitor-export --input examples/catalog.json --output-dir outputs/catalog
uv run exhibitor-export --input examples/technology-and-agriculture.json --output-dir outputs/combined
uv run pytest
uv run ruff check .

Each local run writes companies.json, companies.csv, summary.json, and ATTRIBUTION.txt. Failed/partial runs return exit code 1 and still write their diagnostic report. Invalid input returns exit code 2.

To run the actual Apify SDK entry point locally:

cp examples/agrotica.json INPUT.json
uv run python -m exhibitor_actor

INPUT.json, local storage, outputs and credentials are ignored by Git. Do not commit credentials.

Docker and deployment

docker build -t helexpo-exhibitor-export:local .
docker run --rm \
  -v "$PWD/examples/agrotica.json:/app/INPUT.json:ro" \
  helexpo-exhibitor-export:local

The .actor directory contains Actor metadata plus input, output and Dataset view definitions. Deploying, setting a price and publishing are separate steps requiring an Apify account. Nothing here automatically creates or publishes a cloud Actor.