Domain‑Specific Databases and Aviasales Mailings
summary
Describes building a domain-specific data store for Aviasales to support notification emails when matching tickets appear. Uses a custom non-SQL database to index large JSON search-result documents (200–800 KB) produced into a queue, and evaluates an approach to efficiently match millions of subscriptions—each composed of multiple predicates—without overbuilding or long development time; the described solution reached production within months and paid back quickly.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 10:11Problem and requirements
- 14:45Design pitfalls
- 15:28Logical expressions structure
- 26:50System optimizations and architecture
- 33:10Results and business impact
- 35:46Q&A
related talks
41mHow to choose the right database for your application
01h 05mDatabases: Bespoke Tailoring and Perfect Fit
43mDatabase in a SaaS application
47mExtendable by Design: Building Generative AI Apps with Postgres and Vector Storage and Azure AI
26mPersonalizing User Content - The Power of a Vector & Search DB
51mEvolution of the PostgreSQL Storage for Listings at Avito
39mSearch Architecture at Avito
28mUsing a database in a data science project - Lessons learned in production