От перфокарт к AI-агентам: как менялась работа компьютеров с текстом за 70 лет
Summary
Sergey Pichin interviews Yandex 360 executives Jens Kroglov and Nikolai Ermoshenko about the technical evolution of Yandex Documents. They discuss the transition to a proprietary engine, handling XML compatibility, and the complexities of online spreadsheets. The conversation covers text encoding history, Unicode, and the shift from chat-based AI to contextual agents for document editing.
Topics discussed
Introduction to modern document services and text interpretation
Technical architecture and comparison with Microsoft and Google
Yandex 360 integration and product ecosystem
Technical challenges of spreadsheets and large data processing
Fun features like Doom in documents and team size
Release cycles, updates, and the Arcane engine
History of text encoding: Perforated cards to ASCII
Unicode standard and symbol representation
Document formats: Markdown, XML, and OpenXML structure
Format conversion and preserving layout fidelity
AI integration: Grammar checking and agent assistants
Alisa Pro assistant and future UX for AI agents
Real-time collaboration and conflict resolution algorithms
Cross-device consistency and font licensing issues
Market competition: Notion, Microsoft, and B2B segments
Desktop apps, Yandex Cloud, and self-hosted solutions
Future of agents, vibe coding, and open source
Personal AI usage and Markdown preferences
Bugs, conferences, hiring, and closing remarks
Listen ad-free on Castria