PredictLeads Company Enrichment

Enrich a list of company domains with real-time signals from the PredictLeads API across 10 endpoints — company profiles, job openings, news events, technology detections, financing, connections, similar companies, GitHub repos, products, and website evolution. Configurable concurrency makes large lists fast.

This actor enriches a known list of companies. To discover companies by filters (location, size, technology, funding), use the PredictLeads discovery endpoints directly — see the PredictLeads API docs.

Quick Start

1. Set credentials — enter your PredictLeads API Key and API Token.
2. Add companies — paste the domains or PredictLeads company IDs to enrich.
3. Pick modules — select one or more endpoints.
4. Run — click Start and find clean, flat records in the dataset.

Input Reference

Field	Type	Default	Description
apiKey	string	—	PredictLeads API key (secret).
apiToken	string	—	PredictLeads API token (secret).
companyInputs	string[]	—	Company domains or PredictLeads company IDs to enrich. Required.
selectedModules	string[]	—	Endpoint modules to run (see list below).
maxResultsPerCompany	integer	100	Max records per company/module, and your main cost dial (see API Credits). 0 = unlimited.
concurrency	integer	3	Domains processed in parallel (1–10).
requestDelayMs	integer	300	Milliseconds between API requests (min 100).
proxy	object	—	Optional Apify proxy configuration.

Available Modules

Module	Description
company	Base company profile — name, description, location, ticker, parent/subsidiary companies.
job_openings	Active and historical job postings from a company career pages.
news_events	Structured and categorized company news events from news sites, PR sites, and company blogs.
technology_detections	Technologies used by companies from job descriptions, company websites, DNS records, and more.
financing_events	Funding companies receive derived from news events.
connections	Customer/vendor, partners, and investor relationships between companies.
similar_companies	Company lookalikes, ranked and with reason.
website_evolution	Tracked subpages of a company's website with content in markdown.
github_repositories	Open-source repos linked to the company.
products	Products and services offered by the company.

Example Input

{
  "apiKey": "your-api-key",
  "apiToken": "your-api-token",
  "companyInputs": ["stripe.com", "shopify.com"],
  "selectedModules": ["company", "job_openings", "news_events", "technology_detections"],
  "maxResultsPerCompany": 100,
  "concurrency": 3
}

Example Output

{
  "module_name": "job_openings",
  "fetched_at": "2026-07-21T12:00:00.000Z",
  "input_identifier": "stripe.com",
  "job_opening_id": "4239672f-ec39-4031-aa72-c93a04446f66",
  "company_domain": "stripe.com",
  "company_name": "Stripe, Inc.",
  "company_ticker": null,
  "title": "Staff Data Engineer",
  "url": "https://stripe.com/jobs/listing/staff-data-engineer",
  "location": "San Francisco, California, United States, Northern America, Americas",
  "seniority": "mid_senior",
  "categories": "engineering, software_development",
  "posted_at": null,
  "first_seen_at": "2026-07-18T10:00:00Z",
  "last_seen_at": "2026-07-19T10:06:54Z",
  "onet_data": { "code": "15-1254.00", "family": "Computer and Mathematical", "occupation_name": "Web Developers" },
  "location_data": [
    { "city": "San Francisco", "state": "California", "country": "United States", "region": "Northern America", "continent": "Americas", "fuzzy_match": false }
  ],
  "raw_extra": {}
}

Each record includes module_name, fetched_at, input_identifier, mapped fields, and raw_extra for any unmapped PredictLeads attributes.

Output & Dataset Naming

Every record is written to two places:

• The run's default dataset — shown in the Console's Output tab and used by standard exports (CSV/JSON/Excel). Records from all selected modules land here together; each carries a module_name field so you can tell them apart.
• A per-module named dataset for a clean split by type:

predictleads-job-openings
predictleads-news-events
predictleads-technology-detections
...

Performance Tips

• Concurrency: Set concurrency to 5–10 for runs with many domains. Each domain processes modules sequentially to avoid rate limits, but multiple domains run in parallel.
• Request delay: The default 300ms is conservative. If you have a high-quota plan, try 100–200ms.
• Max results: Set a lower maxResultsPerCompany (e.g. 25) for initial exploration, then increase for production.
• Timeout: Individual domains time out after 5 minutes. The actor gracefully shuts down 60 seconds before the Apify actor timeout.

API Credits

This actor calls the PredictLeads API with your credentials, so runs consume your PredictLeads API credits (Apify platform usage is billed separately). The actor is built to spend as few as possible by requesting exactly what you ask for in a single page where it can.

How billing works:

• Most endpoints — job openings, news events, technology detections, financing events, connections, website evolution, GitHub repositories, products, and the company profile — are billed 1 credit per API request (page). The actor sets the page size to your maxResultsPerCompany (page size max 1000), so requesting up to 1000 records costs 1 credit per company per module. Asking for more (or 0 for unlimited) pages again — roughly 1 extra credit per additional 1000 records.
• Similar Companies is billed differently: 1 credit per lookalike company returned (not per page). Requesting 10 similar companies costs 10 credits. The endpoint returns at most 50 (default 20).

Estimating a run (credits ≈ per company × number of companies × selected modules):

Module	Credits per company
Company profile & the other 8 normal endpoints	~1 per module (≈ ceil(records ÷ 1000))
Similar Companies	1 per company returned (e.g. maxResultsPerCompany = 10 → 10)

Example: 100 domains, running job_openings (≤1000 each) + similar_companies (10 each) ≈ 100 credits for job openings + 1,000 credits for similar companies ≈ 1,100 credits. Lower maxResultsPerCompany — especially for Similar Companies — to spend less.

Rate Limits and Quota

PredictLeads enforces per-account API quotas and rate limits. The actor retries 429 responses using Retry-After headers and retries transient 5xx errors with exponential backoff. Monitor your PredictLeads dashboard for remaining quota.

Run Summary

After every run, the actor saves a machine-readable RUN_SUMMARY to the default key-value store with total records, errors, skipped modules, duration, and per-module/per-domain breakdowns. Access it via the Apify API or console.

FAQ

Can I run only one module? Yes. Select a single module and one or more company domains.

What happens if a company is not found? The actor logs a warning, marks it as skipped, and continues with the next module or domain.

Why is raw_extra included? PredictLeads may return additional attributes beyond the mapped schema. raw_extra preserves them without breaking the stable output format.

Can I use PredictLeads company IDs instead of domains? Yes. companyInputs accepts both domains and PredictLeads company identifiers.

What if I paste a full URL like https://www.stripe.com/? The actor automatically cleans it to stripe.com.

Does the actor stop if one module or domain fails? No. Errors are logged per-module and summarized at the end while the run continues.

Links

• PredictLeads API Documentation
• Apify Actor Documentation

Changelog

3.0.0

• Added company module (base company profile — descriptions, location, ticker, parent/subsidiary/similar companies).
• Credit-efficient paging: the API page size (limit) now follows maxResultsPerCompany (capped at 1000, or 50 for Similar Companies), so small fetches complete in one request instead of many — fewer PredictLeads credits and less API load. See the new API Credits section.
• Removed discovery modes; this actor is now focused solely on enriching a known list of companies. For company discovery, use the PredictLeads discovery endpoints directly (see the API docs).
• List fields (tags, categories, contract_types, …) are now comma-joined strings for clean single CSV columns.
• Added an output-schema view so exports lead with the most useful columns instead of alphabetical order.
• Records are written to both the default dataset (Console Output tab / exports) and per-module named datasets.

2.0.0

• Added configurable concurrency for parallel domain processing (3–10x faster).
• Added input sanitization (URL cleaning, whitespace trimming, lowercasing).
• Added deduplication of records across pages.
• Added per-domain timeout (5 minutes) and graceful shutdown on SIGTERM/actor timeout.
• Added machine-readable RUN_SUMMARY to key-value store.
• Added structured logging with timing, progress percentage, and summary table.
• Added concurrency, requestDelayMs, and proxy input fields.
• Added connection pooling with keep-alive for reduced latency.
• Added smart pagination (stops early when the reported count is reached).
• Added QuotaExhaustedError (402) handling.
• Added human-readable module labels and grouped sections in input UI.
• Improved null safety: unmapped fields return null instead of undefined.
• Refactored to module registry pattern, centralized constants, and typed error hierarchy.

1.0.0

• Initial release with 17 PredictLeads endpoints, company and discovery modes.