Using AI‑assisted coding, Qunar’s team transformed a 150,000‑line high‑concurrency core system, tackling system‑level uncertainties through a harness‑and‑loop framework, achieving roughly 70% faster delivery, 70% lower latency, 45% resource savings, and a fault‑free launch without QA intervention.
The Genos team introduces Gengram, a 20‑million‑parameter plug‑in that stores 1‑6‑mer embeddings in a hash memory, uses local window aggregation and gated writing, and delivers up to 22.6% performance gains across multiple genomic tasks while accelerating training.
This case study details NetEase Cloud Music's full‑link baseline governance initiative, outlining the challenges of massive data pipelines, the metrics used to measure success, the three‑pronged action plan (infrastructure, task optimization, and standards), and the resulting improvements in availability, resource utilization, and monitoring accuracy.
This article details NetEase Cloud Music's cross‑platform network library practice, describing the challenges of duplicated implementations, the design of a unified Cronet‑based solution, upgrade strategies, pitfalls, and real‑world performance gains of up to 40% when enabling QUIC.
The talk explains how Youzan’s rapidly expanding 1,000‑engineer tech group can regain speed and quality by applying three classic management principles—flow value, hole theory, and leverage points—to focus early on a small core of 50 high‑capability staff through training, mentorship, metrics, dedicated support teams, delegated authority, and clear architect versus technical‑leader roles.
