FAQ from KWatch.io
What is KWatch.io?
A real-time AI listening layer for developers, marketers, and product teams—designed to detect, classify, and contextualize brand-related signals across Reddit, Twitter (X), and Hacker News—without requiring engineering overhead or API key management.
How to use KWatch.io?
No code required. Add keywords → choose sources → set alerts → review AI-enriched insights. All logic—including deduplication, source prioritization, and semantic relevance filtering—is handled automatically in the background.
Can I stop my paid subscription anytime?
Yes—no lock-in, no penalties. Subscriptions are month-to-month and prorated down to the hour. Cancel anytime; access remains active until the end of your current billing period.
Do you monitor LinkedIn?
Not currently—but we're actively evaluating official API pathways. Suggest it directly: [email protected]. Priority is given to requests backed by team subscriptions.
Do you monitor Facebook?
We don’t ingest public Facebook pages or groups at this time due to platform restrictions and signal noise. We focus on high-intent, open-source communities where authentic, unfiltered opinions thrive.
Do you monitor Twitter?
Yes—with precision. Our integration respects X’s latest API policies and is available on all paid plans. We filter out bots, retweets without commentary, and low-engagement noise to surface only meaningful human discussion.
Do you monitor keywords in real time?
Yes—average detection latency is under 90 seconds from post publication. We poll continuously and prioritize freshness without sacrificing accuracy.
How will you inform me when you detect a keyword?
You’ll receive a clean, digestible email with title, source link, excerpt, sentiment score, and confidence level—or push raw JSON to your endpoint via secure webhook with customizable headers and retry logic.
Can I send the alerts to several persons?
Absolutely. Invite teammates to a shared workspace, assign role-based permissions (viewer/editor/admin), and route alerts per user or channel—ideal for cross-functional response workflows.
Which kind of AI analysis do you perform exactly?
We run fine-tuned transformer models trained specifically on tech and startup discourse—scoring sentiment (positive/neutral/negative), detecting sarcasm and irony, identifying intent (question, complaint, recommendation), and highlighting emotional intensity—all calibrated for Reddit’s nuance and Hacker News’ technical rigor.
A question? A suggestion? Please reach out!
We read every message. [email protected]