All products
Product · 18 / 19
Content Analysis
Paused · scripts onlyFor: Strata Communications
Article retrieval and AI classification feeding Strata's competitive-intel pipeline.
CLI scripts only — no deployed UI
No screenshot yet — CLI scripts only — no deployed UI.
What it does
How it works.
Python scripts that query NewsCatcher's CatchAll API using pre-built search terms, retry failed/timed-out queries, download each article's HTML, and ask OpenAI's gpt-5-nano to score whether it is genuinely about the client (e.g. EMASS semiconductor), producing a filtered, confidence-scored article list exported to CSV/XLSX. No UI, API, or deployed service — output is local spreadsheet files inspected by hand.
- 01Polls NewsCatcher's CatchAll async job API and aggregates paginated results
- 02Rotates across multiple NewsCatcher API keys when one is rate-limited/exhausted
- 03Retries only the failed/timed-out queries from a prior run
- 04Downloads article HTML and runs async (100-concurrent) OpenAI relevance scoring
Current status & what's next
Where it stands.
Paused · scripts only
No further phases planned.
Stack
Built with.
Python 3requests/pandas/aiohttp/BeautifulSoupOpenAI API (gpt-5-nano)NewsCatcher CatchAll APILet's build what's next
Want Content Analysis working for you?
Article retrieval and AI classification feeding Strata's competitive-intel pipeline.
Get in touchNext product
ScoreCard 2.0