Building AI that Searches and Sells
summary
LeadGenius builds a data pipeline that scrapes ~55 sources using Scrapy and BeautifulSoup plus browser automation (Selenium), running on an Azure cluster that pulls 7.5 million company records into ~202 GB for about $50 per run. Deduplication uses 30-dimensional feature vectors or a USPS-normalized-address shortcut; dataset yields ~6 million “interesting” multi-person companies. Purchase intent is a probabilistic similarity score escalated to human annotators; outbound email is template-driven, trained on customer inboxes, and classifies replies into eight types with sentiment models.
AI-written summary · about
chapters (7)
- 0:00Introduction and vision
- 1:59Salesperson daily tasks
- 3:22AI search for customers
- 8:15Company matching techniques
- 13:12AI understanding buying intent
- 15:58AI email communication
- 20:16Q&A and closing
related talks
27mDisrupting Lead Generation through Artificial Intelligence
07mInterview with – Arthur Jansen
11mGenesis Startup Competition Pitch: SparkBeyond
28mML & GenAI for Vibe Marketing: Future of Audience Connection
43mImplementing AI Document Processing at Scale: Syngenta
47mAI Pitch: Beyond the Hype
11mThe magnet and the spear: AI for client targeting
25mAugmented Intelligence — The Future of GenAI?