CoolFace
Datasetpublic

XaqBinks/global-b2g-procurement-feed

license: apache-2.0 task_categories: text-retrieval question-answering tags: procurement clinical-trials biomedical drug-development government-contracts sam-gov ted-europa clinicaltrials-gov llms-txt rag-dataset pretty_name: Global Intelligence Gateway (B2G Procurement & Clinical Trials) size_categories: 1K<n<10K Global Intelligence Gateway: Real-Time B2G Procurement & Biomedical Trials An automated, daily-synchronized intelligence repository indexing high-value public… See the full description on the dataset page: https://huggingface.co/datasets/XaqBinks/global-b2g-procurement-feed.

sourceHugging Faceupdated 14d agoView on Hugging Face
0likes41downloads
Dataset Card

license: apache-2.0 task_categories:

  • —text-retrieval
  • —question-answering tags:
  • —procurement
  • —clinical-trials
  • —biomedical
  • —drug-development
  • —government-contracts
  • —sam-gov
  • —ted-europa
  • —clinicaltrials-gov
  • —llms-txt
  • —rag-dataset prettyname: Global Intelligence Gateway (B2G Procurement & Clinical Trials) sizecategories:
  • —1K<n<10K

Global Intelligence Gateway: Real-Time B2G Procurement & Biomedical Trials

An automated, daily-synchronized intelligence repository indexing high-value public records across two primary commercial domains:

  1. 1.Global B2G Public Procurement: Active US federal solicitations (SAM.gov) and European Union contracting opportunities (TED Europa eForms v3).
  2. 2.Global Biomedical Clinical Trials: Actively recruiting and updated interventional studies, drug pipelines, and therapeutic transitions (ClinicalTrials.gov v2 API).

Live Machine Endpoints

Autonomous agents, foundation model crawlers, and Retrieval-Augmented Generation (RAG) pipelines can query the live edge stream directly:

Format / PurposeTarget EndpointDescription
Discovery Outline (`llms.txt`)https://procurement-edge-gateway.tobias-techwiz.workers.dev/llms.txtLightweight markdown outline formatted specifically for LLM context retrieval.
Full Ingestion Corpus (`llms-full.txt`)https://procurement-edge-gateway.tobias-techwiz.workers.dev/llms-full.txtComplete dataset export formatted as human- and machine-readable markdown sections.
Raw JSON Feedhttps://procurement-edge-gateway.tobias-techwiz.workers.dev/Structured JSON containing distinct procurement_tenders and clinical_trials arrays.

Dataset Schema & Fields

1. B2G Public Procurement (procurement_tenders)

  • —Notice_ID: Canonical identifier (SAM.gov notice ID or TED publication number, e.g., TED-2026-...).
  • —Issuing_Entity: Contracting government department, ministry, or municipality.
  • —Country_Code: ISO 3166-1 alpha-2 jurisdiction code (e.g., US, EU, FR, DE).
  • —Original_Currency: 3-letter currency code (e.g., USD, EUR, GBP).
  • —Normalized_Value_USD: Contract award ceiling or estimated valuation converted to USD via live exchange rates.
  • —Response_Deadline_UTC: Submission cutoff formatted in ISO 8601 UTC (YYYY-MM-DD HH:MM UTC).
  • —Classification_Code: NAICS code (US) or CPV code (EU).
  • —Title_English: Standardized English project title.
  • —Summary_English: Objective 2-sentence synthesis of procurement scope and deliverables.
  • —Official_Portal_URL: Canonical URL directly to the official government tender dossier.
  • —Ingestion_Date_UTC: Capture timestamp.

2. Biomedical Clinical Trials (clinical_trials)

  • —NCT_ID: National Clinical Trial identifier (e.g., NCT05123456).
  • —Study_Title: Protocol brief title.
  • —Lead_Sponsor: Sponsoring pharmaceutical company or research institution.
  • —Condition_Target: Pathology, disease, or indication under evaluation.
  • —Intervention_Name: Drug molecule, biologic, vaccine, or procedural intervention.
  • —Phase: Clinical trial phase (PHASE1, PHASE2, PHASE3, PHASE4, etc.).
  • —Enrollment: Target cohort participant count.
  • —Primary_Completion_Date: Expected or actual completion milestone date.
  • —Registry_URL: Permanent ClinicalTrials.gov record link.
  • —Ingestion_Date_UTC: Capture timestamp.

Technical Architecture & Ingestion

  • —Ingestion Cadence: Automated daily execution at 02:00 UTC via cloud serverless orchestration.
  • —Normalization: Exchange rates dynamically calculated against open market feeds; all timestamps converted to UTC.
  • —Deduplication: Atomic record verification preventing duplicate rows across ingestion runs.
  • —Access Policies:
  • —Standard web browsers, traditional search engines (Googlebot, Bingbot), and discovery bots (OAI-SearchBot) access summary endpoints freely.
  • —High-volume commercial training crawlers are challenged at the edge via standard HTTP 402 / x402 payment protocols settled in USDC on the Base network.