How AI Fuels BrokerChooser’s Growth
summary
BrokerChooser translates its template-driven broker pages into 70 languages to scale from 100 brokers to ~7,000 localized pages, producing about 1 million pages in January and now generating roughly half the site’s traffic; translations launch in three weeks, use GPT/foundation models and an LLM-as-judge evaluation loop to pick the best model. An assistant called Nuri raises conversion for users who use it but sees low adoption so is paused. Data collection is experimental—using Perplexity, Crawl for AI and custom pipelines—aiming to cover 3,000 brokers.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:04Impact over AI tech
- 3:12BrokerChooser mission
- 7:13AI translation impact
- 13:44AI assistant Nuri conversion
- 18:42AI data collection challenges
- 21:54Key learnings
- 24:04Q&A
related talks
27mDisrupting Lead Generation through Artificial Intelligence
37mLeveraging Generative AI for Enhanced E-commerce
43mBreaking the language barrier: how do we quickly add multilanguage support in our AI application?
21mBuilding AI that Searches and Sells
24mState of Startups and AI 2025
14mThe Retail Revolution: Real-World Generative AI Applications
24mHow Bloomberg Uses AI to Help Speed up Over-the-Counter Trading
14mAI That Moves Business. Cloud That Scales