YUQIAN Y.

When systems break under real-world constraints — unstable networks, missing documentation, mixed hardware — I step in to diagnose and fix them. I specialize in: • Systems integration (IoT, edge devices, multi-protocol environments) • Real-time streaming (WebRTC / RTSP / low-latency pipelines) • Performance & storage engineering (high-throughput, large-scale systems) • Distributed systems & Go backend (control-plane, metadata services, cluster coordination) Selected work: • Airport perimeter security platform Integrated fiber optic vibration sensors, IP cameras, alarms, and lighting into a unified real-time alert platform. → Received hands-on training from Fibersonics engineers on sensor physics and signal analysis — applied this to configure per-zone detection parameters, achieving near-zero missed detections with low false positives across outdoor conditions → Debugged and stabilized WebRTC/RTSP streaming through packet-capture analysis — no documentation available; resolved by reading ZLMediaKit source code and tracing protocol negotiation → Deployed and in active use • Offline mesh-based classroom system Built ESP32-C6 mesh network for student response collection in zero-WiFi environments. → Power-optimized to ~1 month battery life (4000mAh) in testing → Designed full stack: embedded firmware + desktop management + cloud analytics • Offline canteen management system Implemented custom TCP server on Android with a Windows C++ Qt client over direct ethernet. → Ensured data integrity for large transfers in a network-restricted environment with no external infrastructure → Deployed and in active use • Metadata control-plane for disaggregated database (major state-owned bank) Core developer of the Go-based metadata/control-plane service that enabled compute-storage separation on a Greenplum/PostgreSQL-based distributed database. → Centralized TB-scale metadata — transactions, sessions, locks, invalidation state — previously trapped inside local compute nodes → Built on FoundationDB (core KV backend), etcd (cluster coordination, leader election), and a custom RPC layer → Implemented session lifecycle, distributed locking, deadlock detection, and invalidation message pipeline • Disaggregated storage system (China Construction Bank) Core developer of the disaggregated storage module managing thousands of PBs across multi-cloud object storage (AWS S3, Alibaba Cloud, Huawei Cloud, Tencent Cloud) behind a ~1TB local cache. → Designed query-driven prefetching and segmented caching of hot data → Cold start ~20x slower than SSD; warm cache performance on par with SSD • Adaptive compression engine (YMatrix time-series database) Designed automode system that dynamically detects data characteristics across stream segments and selects optimal encoding chains (zigzag, simple8b, deltadelta, Gorilla2, lz4/zstd). → Deployed in production — patented encoding chain design Prior to the above, I spent 6 years as Lead Engineer at FreeWheel (a US media-tech company, acquired by Comcast), building Kubernetes-based microservices platforms and cutting fault localization time from over a day to under a minute using Istio service mesh and distributed tracing. Before that, I was a Technical Expert at Alibaba Cloud, where I extended a database proxy to support the PostgreSQL protocol (read/write splitting, failover reconnection) and built a stability engineering system for RDS PostgreSQL that kept the service running through 2021 Double 11 — one of the world's highest-traffic e-commerce events. I've diagnosed production issues across Go, C++, and distributed systems — including a Golang memory leak traced through Prometheus metrics to a misconfigured HTTP client, and C++ performance bottlenecks in deployed systems. My approach: correlate monitoring data with business operations to narrow scope before touching the code. I've published source-level analysis of Prometheus internals on InfoQ. Domain experience with counter-drone detection systems — participated in field testing and evaluation of RF detection and optical tracking platforms, with firsthand understanding of operational scenarios, detection requirements, and alert workflow needs. If you're dealing with: • unstable integrations • streaming issues (WebRTC / RTSP) • hardware + software systems not working together • performance problems in production —I can help you find and fix the root cause. Tech stack: Systems Integration Engineer | Edge AI, Video, Embedded & Distributed Systems