Australian Government Procurement Re-Tender Ranking

Rank Australian Government procurement candidates by historically observed recurrence patterns using a frozen, leakage-safe model.

Use it for supplier intelligence, tender monitoring, procurement research, API pipelines, and AI-agent workflows that need a machine-readable priority score rather than another tender list.

Why use this instead of an AusTender scraper

An AusTender scraper collects notices and contracts. This Actor does not duplicate that job: it takes candidate procurement episodes and adds a historical recurrence-ranking layer so downstream systems can prioritize which candidates deserve attention first. Use a scraper to build the candidate set; use this Actor to rank it.

Fast start

{
  "candidates": [
    {
      "agency": "Department of Education",
      "category": "86000000",
      "scoreDate": "2026-09-25",
      "publishDate": "2026-08-25"
    }
  ],
  "maxResults": 25
}

What you get

• ranked procurement candidates
• foresightScore from 0-100
• agency, category and scoring date retained in every result
• frozen model version and historical evidence lineage

Pay only for delivered results

The primary Pay Per Event charge is foresight-result at US$0.03 per delivered scored result. Dataset delivery and the custom PPE event are coupled so the Actor stops when the event budget cannot accept another paid result.

Evidence boundary

foresightScore is a relative recurrence ranking score, not a calibrated probability and not a prediction that a tender will definitely reappear. The production model is frozen from the validated V19 leakage-safe ranking work and discloses its historical cutoff and model version in output evidence.

Machine-ready design

The Actor is designed for Apify API calls, scheduled procurement pipelines, and AI-agent execution. It requires no external API key from the caller and returns structured dataset rows suitable for downstream filtering, alerts, dashboards, or model inputs.

AI-agent example

Ask an agent: Rank these Australian Government procurement candidates by historical re-tender recurrence signals and keep the score limitations visible in every result.

Input notes

Each candidate requires agency, category, and either scoreDate or endDate. publishDate is optional and is used to derive episode age. Send multiple candidates in one run to obtain a relative priority ordering.