Qunar Tech Salon is a learning and exchange platform for Qunar engineers and industry peers. We share cutting-edge technology trends and topics, providing a free platform for mid-to-senior technical professionals to exchange and learn.
This article details how a leading OTA transformed its hotel review pipeline with multimodal large‑language models, real‑time event‑driven architecture, and automated static‑info correction, achieving sub‑second moderation, 99.6% accuracy, and measurable cost and user‑experience gains.
This article examines the challenges of pixel‑perfect UI reconstruction and semantic code generation for consumer‑facing front‑end projects, and presents a hybrid rule‑algorithm and large‑language‑model architecture that streamlines design‑to‑code (D2C) and requirement‑to‑code (P2C) workflows, integrates IDE plugins, and dramatically improves developer productivity.
The article explains why hotel booking platforms must implement a price‑cache layer, detailing performance bottlenecks, traffic spikes, and data freshness challenges, and describes a split real‑time and offline architecture with dual‑update strategies, cache‑freshness logic, and high‑availability mechanisms to ensure fast, reliable pricing.
This article details the design and implementation of a visual canvas‑based marketing configuration system that replaces Apollo, introduces a custom Reactor‑powered workflow engine, outlines component ecosystems, solves concurrency challenges, and demonstrates significant efficiency, scalability, and reliability gains after deployment.
The article walks through a real‑world OOM incident in a high‑traffic hotel information service, detailing the root‑cause analysis of memory exhaustion, JVM heap configuration, GC behavior, heap‑dump inspection, and the concrete optimizations applied to prevent similar failures.
This article examines the limitations of traditional threshold‑based database inspections, introduces AI‑driven anomaly detection techniques such as DoubleRollingAggregate, SeasonalAD, and LevelShiftAD, and details practical implementations, tuning strategies, and real‑world use cases for MySQL and Redis monitoring.
This article presents practical, step‑by‑step techniques for refining prompts used in large language model applications—covering intent detection, context enrichment, instruction compliance, model capability activation, and structural design—to dramatically improve accuracy, reduce hallucinations, and boost overall AI system reliability.
This article analyzes the shortcomings of the legacy database monitoring system, explains the transition from a push‑based to a pull‑based architecture, outlines comprehensive metric collection, intelligent alert strategies, and self‑healing mechanisms, and showcases the performance improvements achieved with the new solution.
An in‑depth exploration of a Multi‑Agent AI system that automates SDLC white‑box vulnerability management, detailing industry‑standard processes, the system’s architecture, specialized agents, prompt engineering, tool integration, and real‑world results that boost audit efficiency and accuracy while enabling true security left‑shift.
This article details how Quark’s data platform consolidated multiple analytics engines into a unified StarRocks‑based OLAP solution, covering business background, engine selection, architecture redesign, performance tuning, operational practices, and future plans for scalability and reliability.
