This example builds a news monitoring system that searches for recent articles on specific topics, filters by date and source, and collects results for processing.
Complete example
# Search for AI news from the past 24 hours
curl -s "https://search-api.andisearch.com/api/v1/search?q=artificial+intelligence&intent=news&dateRange=24h&limit=10" \
-H "x-api-key: $ANDI_API_KEY" | jq '.results[] | {title, link, date, source}'
How it works
intent=news forces news-specific results rather than general web searchdateRange=24h limits results to the past 24 hours- News array — when the intent is
news, results may appear in the news array alongside results
- Multiple topics — loop through topics to monitor several areas at once
Variations
Restrict to trusted sources
Limit results to specific publications:
tech_sources = [
"arstechnica.com",
"wired.com",
"technologyreview.com",
"theverge.com",
]
articles = search_news(
"artificial intelligence",
date_range="week",
domains=tech_sources,
)
Exclude aggregators
Remove noisy domains from results:
response = requests.get(
"https://search-api.andisearch.com/api/v1/search",
params={
"q": "AI startups funding",
"intent": "news",
"dateRange": "week",
"excludeDomains": "reddit.com,medium.com",
},
headers={"x-api-key": api_key},
)
Custom date range
Search a specific time window:
response = requests.get(
"https://search-api.andisearch.com/api/v1/search",
params={
"q": "product launch",
"intent": "news",
"dateFrom": "2025-03-01",
"dateTo": "2025-03-15",
},
headers={"x-api-key": api_key},
)
Scheduled monitoring
Run searches on a schedule using cron or a task scheduler:
import time
def monitor_loop(topics: list[str], interval_seconds: int = 3600):
"""Check for new articles every interval."""
while True:
for topic in topics:
articles = search_news(topic, date_range="24h")
if articles:
print(f"[{datetime.now()}] {topic}: {len(articles)} new articles")
# Process articles: send alerts, save to database, etc.
time.sleep(interval_seconds)
When running scheduled searches, respect your rate limits. Space requests out and use the X-RateLimit-Remaining header to monitor usage.
Next steps
