Dataset Viewer

The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.

Python Code:

import sqlite3
from datasets import load_dataset
import pandas as pd

# Method 1: Load via Hugging Face datasets library
dataset = load_dataset("keyword-barrage/enterprise-sku-depletion-telemetry")
print("Hub Sample Record:", dataset["train"][0])

# Method 2: Load zero-latency local pre-indexed SQLite deliverable
# Optimized for rapid querying of SKU velocity and price elasticity
conn = sqlite3.connect("enterprise_feed.sqlite")
df = pd.read_sql_query("""
    SELECT sku_id, price_usd, stock_level, depletion_velocity 
    FROM sku_stream 
    WHERE timestamp > datetime('now', '-24 hours')
    ORDER BY depletion_velocity DESC LIMIT 10
""", conn)
print(df)

Dataset Card:

Dataset Summary

This dataset provides a production-grade, high-frequency telemetry feed for Vivino.com, engineered to bypass AWS CloudFront/WAF Layer 7 challenges. Designed for hedge funds, quantitative market analysts, and e-commerce leaders, this feed eliminates the "scraping tax"—the high failure rates and maintenance overhead associated with residential proxy pools and headless browser emulation.

Under our 1,500 USDT/month SLA Retainer, we deliver a fully-managed, 99.9% uptime data pipeline. We handle the protocol-level handshake, TLS/JA4 fingerprinting, and session persistence, ensuring your downstream RAG pipelines and ML models receive clean, schema-verified intelligence without the noise of bot-mitigation triggers.

Enterprise Deliverables Matrix

We provide four distinct formats optimized for diverse data architectures:

  1. Pre-indexed SQLite (.sqlite): Instant local querying with pre-built B-Tree indices on sku_id and timestamp.
  2. Snappy Parquet: Columnar storage optimized for high-throughput ingestion into Snowflake, Google BigQuery, or AWS Athena.
  3. Offline DataGrid (.html): A self-contained, interactive UI for non-technical business stakeholders to visualize SKU velocity and price shifts.
  4. CDC JSONL Stream: Change Data Capture format designed for event-driven architectures and real-time vector database updates.

Dataset Schema

Column Type Nullable Description
sku_id STRING No Unique identifier for the vintage/product.
timestamp DATETIME No UTC precision timestamp of extraction.
price_usd FLOAT No Real-time merchant listing price.
stock_level INT Yes Current inventory delta or availability flag.
depletion_velocity FLOAT No Calculated rate of inventory movement (units/hour).
waf_status STRING No Telemetry verification of the WAF bypass success.
tls_ja4_sig STRING No TLS fingerprinting signature for session parity.

Sample Record (JSON):

{
  "sku_id": "VIV-98210-2018",
  "timestamp": "2026-09-22T14:30:00Z",
  "price_usd": 89.99,
  "stock_level": 42,
  "depletion_velocity": 0.15,
  "waf_status": "BYPASS_VERIFIED",
  "tls_ja4_sig": "t13d15h2h3_8932_2319"
}

SLA & Compliance

  • Availability: 99.9% SLA on data availability.
  • Support: 24/7 monitoring of upstream WAF updates; immediate pipeline recalibration.
  • Compliance: All datasets are collected via protocol-level parity, ensuring zero impact on target site infrastructure or load.
  • Engagement: Minimum 1,500 USDT/month retainer. Pay-on-delivery guarantee: verify the sample data quality before payment is settled.

https://www.keywordbarrage.com/scrapers/vivino-scraper.html

#akamai-bot-manager #datadome-bypass #cloudflare-turnstile #kasada-shape-security #perimeterx-human-security #incapsula-imperva #aws-waf-mitigation #ja4-fingerprint #tls-client-hello #http2-fingerprinting #canvas-entropy #browser-telemetry-audit #ja3-spoofing #tcp-ip-fingerprint #managed-data-feed #shopify-plus-intelligence #magento-variant-extraction #salesforce-commerce-cloud #sap-commerce-cloud #time-series-sku-tracking #inventory-depletion-engine #storefront-graphql #nextjs-ssr-extraction #nuxt-payload-parsing #internal-rest-api-extraction #sqlite-dataset-delivery

📦 Turnkey 4-in-1 Data Delivery & Standalone Offline Viewers

Every dataset includes a pre-indexed SQLite DB, structured JSON, flat CSV, and our proprietary Zero-Dependency Interactive HTML DataGrid. Double-click to search, sort, and filter tens of thousands of records offline—zero SQL, zero SaaS subscriptions, and zero API keys required.

Live Interactive Deliverable Demos (No Setup Required):

Downloads last month
32