How Duolingo Rebuilt Its AI Tutor for 100 Million Learners

Episode notes

In this episode, Lucas and Luna explore how Duolingo rebuilt its AI-powered language tutor, Birdbrain, to personalize learning for over 100 million monthly active users. They dive into the technical challenges of scaling real-time adaptive lesson generation, the decision to move from rule-based systems to transformer models, and how the team balanced engagement with educational efficacy. Specific numbers include a 40% increase in user retention and a 15% improvement in learning outcomes after the rebuild. The conversation also touches on the trade-offs between server-side and on-device inference, and the surprising insight that simpler models often outperformed complex ones for certain tasks. A must-listen for anyone interested in AI at scale, edtech infrastructure, or engineering orgs that prioritize experimentation. #Duolingo #AI #EdTech #MachineLearning #Transformers #Personalization #Scalability #Birdbrain #Retention #LanguageLearning #Business #Technology #Engineering #Architecture #Podcast #FexingoBusiness #BusinessPodcast #CTO Keep every episode free: buymeacoffee.com/fexingo

Related episodes across the Index

Other episodes covering the same guests and topics, from across The B2B Podcast Index.

• The Future of Tech - Key Themes for 2026 and BeyondThe B2B Podcast · on Transformer models82 / 100
• Finding Agency With AI In A World That Won’t Slow Down with Andy SitisonMEDIASCAPE: Insights From Digital Changemakers · on Transformer models81 / 100
• How Losing Her Best Friend Led Eugenia Kuyda to Build the World's First AI CompanionVentures from The Valley · on Transformer models76 / 100
• IBM’s AI Bet: Rebuilding Software Development with IBM BobThe Digital Leader Show · on Transformer models76 / 100
• S1, EP17, 2020 Entrepreneurs Club: INDIA SPECIAL & Limechat: Serial entrepreneur, Aniket Bajpai, and launching a Titan VC backed e-commerce solution in the fastest growth tech hub on earthThe Talent Success Podcast · on Transformer models59 / 100
• Why AI Adoption Depends on Hiring the Next Generation of TalentInsuring Cyber Podcast · on Transformer models48 / 100

• How Airbnb Rebuilt Its Search for 100 Million Listings72 / 100
• How GitHub Migrated 100 Million Repositories to a New Storage Engine76 / 100
• How Dropbox Rebuilt Its Sync Engine for 700 Million Users85 / 100
• How Skyscanner Migrated 300 Microservices to Event-Driven Architecture85 / 100
• How Palantir Rebuilt Its Foundry Ontology for Government AI Deployments85 / 100