Revenue Search: Inside Bittensor Subnet Session with Giga from Desearch: Subnet 22
29 snips
Sep 14, 2025 Explore how Subnet 22 transforms LLMs into real-time operators by integrating fresh data sources like web, Twitter, and Reddit. Giga reveals the essential role of real-time search for AI agents and their unique ability to act as 24/7 sales representatives. Learn about the innovative miner workflow that retrieves and ranks data, along with insights into multi-source searching. Plus, Giga discusses the exciting future for startups, pricing strategies, and the value of persistent agents in driving engagement and traffic.
AI Snips
Chapters
Transcript
Episode notes
Real-Time Data Is The Missing LLM Layer
- LLMs are powerful but blind without fresh, real-time data from the web and social media.
- Subnet 22 provides an open real-time search layer to feed LLMs with up-to-the-minute signals.
Pivot From Search App To Data Layer
- Subnet 22 pivoted from a consumer search app to a data layer focused on relevance and freshness for agents.
- Miners train models to retrieve, scrape and rank contextually relevant, real-time information for API queries.
Social Conversations Drive High-Value Signals
- Twitter, Reddit and web conversations are the highest-value real-time sources for crypto and project monitoring.
- Deep search can analyze tens of thousands of tweets to surface sentiment, pros/cons and actionable insights.