Radaric

Product Introduction

1. Definition: Radaric is an AI-powered content curation and aggregation platform (SaaS). It functions as an intelligent feed reader that ingests content from diverse sources including RSS, Telegram (including private channels), YouTube (via transcript analysis), X (Twitter), Substack, Bluesky, Mastodon, Product Hunt, GitHub, and arXiv.
2. Core Value Proposition: It exists to solve information overload by using a personalized AI filter that learns from a user's explicit feedback (approve/reject/explain) to scale their unique editorial judgment. Its secondary proposition is cross-source story deduplication, merging repetitive coverage of the same event into a single, source-linked story to save time.

Main Features

1. Multi-Source Content Ingestion: Radaric connects to over ten content source types. Technically, it uses official APIs (Twitter/X, YouTube, GitHub), client-based parsing (Telegram), and standard protocols (RSS, OPML import). A key differentiator is its ability to access private Telegram channels and read YouTube video transcripts for content evaluation, which most RSS readers cannot do.
2. Personalized, Learning AI Filter: The core AI feature operates on a user-defined prompt ("what matters to you"). It evaluates each incoming post, assigning a confidence score and a reason. Crucially, it employs a feedback loop: every user approval, rejection, or explanatory note trains the model, continuously refining the filter to match the user's taste more accurately than a static prompt.
3. Cross-Source Story Aggregation (Deduplication): This feature uses natural language processing (NLP) to compare posts across all connected feeds. It analyzes titles, content, named entities, and timestamps. When a threshold of similarity is met, it groups posts into one "story," presenting consolidated facts while preserving links to every original source, even across different languages.

Problems Solved

1. Pain Point: "Content Overload and Repetition." Professionals waste hours daily skimming hundreds of near-identical items from multiple feeds and platforms to find genuinely unique, relevant information.
2. Target Audience: Primary personas include Newsletter Editors, Content Curators, Researchers, Industry Analysts, and Knowledge Workers (e.g., developers, designers) who must monitor numerous niche sources for timely, high-signal information.
3. Use Cases: A) A journalist monitoring breaking news across X, Telegram, and YouTube channels, needing to avoid duplicate reports. B) A developer tracking updates on GitHub, arXiv, and tech blogs, wanting only relevant project/news. C) A marketing manager curating industry trends from Substack, RSS, and private Telegram groups.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):
   • Pros: Uniquely combines multi-source ingestion (especially Telegram/YouTube) with a personalized learning AI. The story aggregator is a significant time-saver. Direct developer access for early users is a strong trust signal.
   • Cons: No free tier, which limits casual testing. Twitter/X parsing is locked behind the high-cost Pro plan ($59/mo). As a new, indie tool, long-term reliability and feature development pace are unproven compared to established incumbents.
2. Key Alternatives & Differentiation:
   • Feedly/Inoreader: These are primarily RSS/Atom feed readers. Their AI features (e.g., Feedly Leo) are generic, not personalized per user. They lack native support for Telegram, YouTube transcript reading, and automated cross-source story deduplication. Radaric is more expensive than basic plans but offers a fundamentally different, personalized AI agent.
   • Manual Research Assistant (Human): A human VA can monitor any source and learn preferences but costs ~$2000+/month and operates slowly. Radaric automates this at a fraction of the cost ($9-$59/mo) with near-instant processing, though it may lack nuanced human judgment for extremely complex topics.

Frequently Asked Questions (FAQ)

1. How does Radaric's AI differ from Feedly's AI? Feedly's AI (Leo) applies the same set of pre-trained filters to all users (e.g., "AI Security" or "VC Funding"). Radaric's AI is trained by you; it starts from your custom prompt and evolves based solely on your approval and rejection decisions, creating a unique filter that reflects your personal editorial taste.
2. Is it safe to connect my private Telegram channels to Radaric? According to the developer, the connection is read-only. Radaric only retrieves messages and cannot send, modify, or delete any data in your Telegram account. User credentials are stored in an encrypted format.
3. What happens to posts the AI filter rejects? Rejected posts are not deleted. They remain accessible within the interface, one click away. You can review them and see the AI's stated reason for rejection, which allows for transparency and manual correction that further trains the filter.
4. Can Radaric truly replace my current RSS reader? It can replace the filtering and aggregation function for users who monitor diverse sources (beyond RSS). However, users who require complex RSS-specific organizational features (folders, rules, tags) or who only follow traditional RSS feeds may find Radaric's approach different. Its strength is unification and intelligent reduction, not just RSS organization.
5. Who is behind Radaric and what is their credibility? Radaric is built and operated by Kirill Oleinichenko, the editor of "Zhurnalus," a weekly design digest published for over 8 years (500+ issues). The product was built to solve his own professional curation pain points and has been in his daily use since November 2025, lending operational credibility.