Quantitative hedge fund building ultra-low-latency trading systems
JQ Investments operates a quantitative trading platform built on C/C++, Python, Go, and kernel-level networking (RDMA, DPDK, NVMe) — a stack optimized for microsecond-scale latency rather than general-purpose cloud infrastructure. The active project list reveals deep infrastructure work: compiler optimization, custom storage for petabyte datasets, and high-frequency trading system design. Pain points cluster heavily around latency reduction and training throughput, suggesting the firm is scaling both live trading and research workloads simultaneously, with engineering hiring weighted toward performance-critical roles.
JQ Investments is a quantitative hedge fund founded in 2012 and headquartered in Shanghai. The firm trades equities, futures, and derivatives using proprietary research and machine learning models. The team spans ~201–500 employees across engineering, research, finance, and operations, with technical talent drawn from top universities and prior roles in research labs and technology companies. The platform architecture reflects a commitment to microsecond-scale execution: custom kernel modules, GPU acceleration, and specialized storage pipelines handle market data ingestion, signal computation, and order routing. Research and engineering are tightly coupled, organized around collaborative projects that blend statistical methodology with systems optimization.
Primary: C/C++, Python, Go. Infrastructure: Kubernetes, Linux Kernel, RDMA, DPDK, NVMe, GPUDirect. Frameworks: Gin, Flask, Vue, React. Databases: MySQL, PostgreSQL, MongoDB.
High-frequency trading platform design, ultra-low latency system development, pricing/return prediction models, market signal analysis, and customized petabyte-scale storage solutions. Heavy focus on performance optimization across compilers, operating systems, and networking layers.
Other companies in the same industry, closest in size