The Problem
AI content generators produce plausible-sounding text that cites fabricated statistics, outdated prices, and nonexistent sources. The model has no concept of what is true right now, only what was true in its training data. Publishers who rely on AI content without fact-checking face reputational damage when readers find errors, and legal liability when the content influences purchasing decisions. Manual fact-checking does not scale when you are producing 50+ articles per week.
The Scavio Solution
Scavio provides live search data that content pipelines use to verify claims before publication. Before the AI writes about a product price, it queries Scavio for the current listing. Before it cites a statistic, it searches for the source. The live data grounds the content in reality and provides source URLs for every claim. The pipeline produces content that is both efficient to create and defensibly accurate, with an audit trail of every fact-check query.
Before
Before Scavio, AI-generated content contained plausible fabrications that slipped past human review. Publishers discovered errors from reader complaints, and each correction damaged credibility.
After
After Scavio, every factual claim in AI-generated content is verified against live search data before publication. Content ships with source URLs, and the hallucination rate drops from occasional to near-zero.
Who It Is For
Content teams producing AI-generated articles at scale who need automated fact-checking before publication. Publishers who cannot afford the reputational risk of hallucinated statistics or outdated prices.
Key Benefits
- Live fact-checking for AI-generated claims before publication
- Source URLs attached to every verified claim
- Audit trail of every fact-check query for editorial compliance
- Scales to 50+ articles per week without manual fact-checking
- Current pricing data prevents publishing stale product information
Python Example
import requests
import json
API_KEY = "your_scavio_api_key"
def verify_claim(claim: str) -> dict:
"""Verify a factual claim against live search data."""
res = requests.post(
"https://api.scavio.dev/api/v1/search",
headers={"x-api-key": API_KEY},
json={"platform": "google", "query": claim},
timeout=15,
)
res.raise_for_status()
data = res.json()
sources = []
for item in data.get("organic", [])[:5]:
sources.append({
"title": item.get("title", ""),
"snippet": item.get("snippet", ""),
"url": item.get("link", ""),
})
return {
"claim": claim,
"sources_found": len(sources),
"sources": sources,
"ai_overview": data.get("ai_overview", {}).get("text", ""),
"verified": len(sources) > 0,
}
def verify_price(product: str, claimed_price: float) -> dict:
"""Verify a product price claim against live marketplace data."""
res = requests.post(
"https://api.scavio.dev/api/v1/search",
headers={"x-api-key": API_KEY},
json={"platform": "amazon", "query": product},
timeout=15,
)
res.raise_for_status()
for item in res.json().get("organic", []):
if item.get("price"):
actual = item["price"]
return {
"product": product,
"claimed_price": claimed_price,
"actual_price": actual,
"accurate": abs(actual - claimed_price) / claimed_price < 0.1,
"source_url": item.get("link", ""),
}
return {"product": product, "claimed_price": claimed_price, "actual_price": None, "verified": False}
# Verify claims before publishing
claim_result = verify_claim("global AI market size 2026")
print(f"Claim verified: {claim_result['verified']} ({claim_result['sources_found']} sources)")
price_result = verify_price("sony wh-1000xm5", 349.99)
if price_result.get("actual_price"):
print(f"Price check: claimed ${price_result['claimed_price']}, actual ${price_result['actual_price']}, accurate: {price_result['accurate']}")JavaScript Example
const API_KEY = "your_scavio_api_key";
async function verifyClaim(claim) {
const res = await fetch("https://api.scavio.dev/api/v1/search", {
method: "POST",
headers: { "x-api-key": API_KEY, "content-type": "application/json" },
body: JSON.stringify({ platform: "google", query: claim }),
});
if (!res.ok) throw new Error(`scavio ${res.status}`);
const data = await res.json();
const sources = (data.organic ?? []).slice(0, 5).map((item) => ({
title: item.title ?? "", snippet: item.snippet ?? "", url: item.link ?? "",
}));
return { claim, sourcesFound: sources.length, sources, verified: sources.length > 0 };
}
async function verifyPrice(product, claimedPrice) {
const res = await fetch("https://api.scavio.dev/api/v1/search", {
method: "POST",
headers: { "x-api-key": API_KEY, "content-type": "application/json" },
body: JSON.stringify({ platform: "amazon", query: product }),
});
if (!res.ok) throw new Error(`scavio ${res.status}`);
for (const item of (await res.json()).organic ?? []) {
if (item.price) return { product, claimedPrice, actualPrice: item.price, accurate: Math.abs(item.price - claimedPrice) / claimedPrice < 0.1, sourceUrl: item.link ?? "" };
}
return { product, claimedPrice, actualPrice: null, verified: false };
}
const result = await verifyClaim("global AI market size 2026");
console.log(`Verified: ${result.verified} (${result.sourcesFound} sources)`);Platforms Used
Web search with knowledge graph, PAA, and AI overviews
Amazon
Product search with prices, ratings, and reviews
Community, posts & threaded comments from any subreddit