NewList your AI tool for $19 $49. Paste a URL and AI drafts the listing.Submit a tool →Already listed? Claim it free.Claim your tool →

Radaric

AI-powered content curation tool that filters and merges news feeds based on personal preferences

$9/moStarting price
Radaric featured image

About Radaric

Radaric is an AI-driven content curation platform designed to help users scale their content consumption by filtering and organizing feeds according to personal taste. The tool connects to multiple sources including RSS, Telegram, Twitter/X, YouTube, Substack, Bluesky, Mastodon, Product Hunt, GitHub, and arXiv. It reads YouTube video transcripts directly and processes content from private Telegram channels. Users define their preferences in plain language, and the AI assigns confidence scores and reasons for each post’s inclusion or exclusion. The system learns from user feedback—approvals, rejections, and saved items—refining its filtering over time. Posts that match user preferences are compiled into a concise feed, reducing the time spent triaging content. When the same news appears across multiple sources, Radaric merges them into a single story while preserving links to all original sources. This aggregation works across different languages and source types, ensuring comprehensive coverage without redundancy.

Key features

  • AI-powered content filtering with confidence scoring
  • Multi-source integration (RSS, Telegram, Twitter/X, YouTube, Substack, Bluesky, Mastodon, Product Hunt, GitHub, arXiv)
  • YouTube transcript reading and processing
  • Private Telegram channel support
  • Story deduplication and merging
  • Personalized learning from user approvals and rejections
  • OPML file import for bulk feed addition
  • Encrypted credential storage

Use cases

  • Reducing time spent filtering through multiple news feeds
  • Consolidating duplicate news stories from different sources
  • Curating personalized content feeds for professional research

Pros

  • Supports multiple content sources including private Telegram channels and YouTube transcripts
  • Learns from user feedback to refine filtering over time
  • Merges duplicate news stories across sources into a single summary
  • Works across different languages and content types
  • Offers automated filtering with manual overrides

Cons

  • No free tier available
  • Twitter/X accounts not included in base plans
  • Requires consistent user feedback to improve filtering accuracy
  • Early access users only get direct developer access temporarily

Frequently asked questions about Radaric

How is Radaric different from Feedly or Inoreader?

Radaric supports multiple sources beyond RSS, including Telegram (including private channels), Twitter/X, YouTube, Substack, Bluesky, Mastodon, Product Hunt, GitHub, and arXiv. Unlike Feedly or Inoreader, its AI learns from individual user feedback—approvals, rejections, and saved items—tailoring the filtering to personal taste rather than applying a generic AI model.

Is my Telegram account safe when using Radaric?

Radaric connects to Telegram in a read-only mode, meaning it only accesses messages without sending, modifying, or deleting anything. User credentials are stored encrypted to ensure security.

Who built Radaric and why?

Radaric was built by Kirill Oleinichenko, the editor of Zhurnalus—a weekly design digest with over 500 issues since 2016. He developed Radaric to address his own challenge of efficiently curating content from hundreds of feeds daily.

Does Radaric support multiple languages for content aggregation?

Yes, Radaric merges and aggregates content across different languages and source types, ensuring comprehensive coverage without redundancy when the same news appears in multiple sources.

Can Radaric read YouTube video transcripts?

Yes, Radaric can read and process YouTube video transcripts directly, allowing it to analyze video content alongside text-based sources.

What happens to posts that are rejected by the AI?

Rejected posts are not deleted; they remain accessible with one click, allowing users to review the AI's reasoning and provide feedback to improve future filtering.

Reviews