Andi Search launches new semantic content engine
In its updated visual feed of search results, Andi is also putting a big priority on clearly citing sources.
Andi Search today is launching a new semantic content engine, as well as better source attribution and a new user experience.
Concepts over keywords. Andi Search is now better at understanding concepts, according to company founder Angela Hoover. The goal is to better highlight content. Being able to rank content semantically helps Andi do that, Hoover said.
- “What makes the web great is the high-quality content that producers spend hours researching, writing and creating. Fair content attribution, however, has been an issue in search from all the way back when Google started doing snippets, and now LLM search opens up a new paradigm for an experience that buries the original content producer even more.
- “Our commitment is to showcase great content and useful information, not steal it. There’s a way to do GenAI search that benefits both the user and producers of media,” Hoover said.
What it looks like. Here’s an example search for [suggest ideas for lowering blood pressure as a 35 year old]:
Visual updates. Andi now uses a “universal image discriminator,” which detects an image’s size and quality, then dynamically resizes it to display in its visual feed. Images are sourced from the webpages in Andi’s search results.
What is Andi. Andi is a startup search engine that was doing conversational, generative question answering before the new Bing and Google Search Generative Experience. It generates results from its index of 30,000 high-quality websites.
Dig deeper. Andi Search adds Generated Content feature.
Why we care. Andi, recognizing the importance of citing its sources, is trying to set itself apart from the generative AI experiences from Google and Bing. While Andi is an alternative search engine with a tiny fraction of Google’s user base and isn’t on your SEO radar, watching the evolution of Andi’s visual, feed-based take on search results is interesting.