Our methodology is designed to help you understand both the story and the limits of the comparison.
We use public RSS feeds and topic searches to find current headlines from Indian and international publishers. We record the headline, publication time, publisher, article URL, and available source metadata. We do not republish full articles.
Related headlines are clustered into a single story using title similarity, shared entities, timing, and URL overlap. Duplicate links are removed. Indian sources are prioritized when they provide relevant reporting, while existing stories are retained during refreshes so the feed does not unexpectedly lose context.
Articles are ordered by relevance and recency, with an effort to include multiple publishers and local reporting. A story may contain fewer sources when there is not enough independent coverage. A high article count should not be read as proof that a claim is true.
An AI-assisted workflow generates a short headline, homepage summary, and overall story brief from article metadata and available text. The prompts ask for neutral language, uncertainty where appropriate, and no source names in the headline. When the AI service is unavailable or rate-limited, a deterministic local fallback is used.
Articles are organized into Left, Centre, and Right sections using provisional publisher-orientation estimates. These describe a broad editorial tendency, not the political position of an individual article. They are not accuracy, quality, or truth scores.
Our separate party hub compares statements from monitored national-party press releases and editor-selected X posts. Each imported item receives an attributed English summary; X posts are displayed through X's official embed and remain the party's own claim, not an Everyside fact check.
Everyside uses publisher-orientation estimates to help readers understand the mix of sources around a story. The label is attached to the publisher, not to the truthfulness of the article or whether the article supports or attacks the subject.
RSS metadata can be incomplete, publishers can change their editorial style, and automated grouping can make mistakes. Breaking news also changes faster than any feed can update. We continuously review errors, but readers should open the original article and check primary documents for important claims.