Business Unit
Cloud & Smart Industries Group (CSIG) is responsible for promoting the company's cloud and industry Internet strategy. CSIG explores the interactions between users and industries to create innovative solutions for smart industries via technological advancements such as cloud, AI, and network security. While driving the digitalization of retail, medical, education, transportation and other industries, CSIG helps companies serve users in smarter ways, building a new ecosystem of intelligent industries that connect users and businesses.
What the Role Entails
About the Role
Tencent Cloud's Big Data team is hiring a Principal Engineer to help define data infrastructure for the AI era. We build data systems that power AI, and we are also bringing AI into the core of storage, compute, and operations, so that our infrastructure can optimize itself and run reliably with less manual effort.
Our work covers two areas. Infra for AI is about building data infrastructure that serves AI workloads. AI for Infra is about using AI to make that infrastructure smarter. At production scale, you will work on hard problems in data throughput, CPU/GPU utilization, and system reliability. You will help set technical direction and architecture, and take ideas from problem definition and research through prototyping to production. We value work that ships, and we also encourage open-source contributions and published research.
We welcome candidates with deep expertise in one or more of the following areas.
What You'll Do
1. Compute: Build distributed compute engines for data and AI workloads. Work on query and operator optimization, CPU/GPU scheduling, workload forecasting, and autoscaling. Help these systems move from static rules to adaptive tuning, and deliver better performance per unit of compute.
2. Storage: Remove data bottlenecks for AI. Design storage architectures for large-scale, multimodal, fast-changing data, covering tiering, hot/cold data detection, data layout, compression, caching, and I/O optimization. Deliver a fast, reliable, and cost-efficient data layer for model training, inference, and AI agents.
3. Data & Search: Combine LLMs, agents, and data systems to create new ways of working with data, such as natural-language queries, text-to-SQL, and automated analytics. Build multimodal data processing, data quality, and metadata capabilities that make enterprise data trustworthy, traceable, and ready for AI. Develop search capabilities, including query understanding, hybrid retrieval, embeddings and reranking, agentic search, and RAG.
4. Self-Managing Systems: Build closed-loop systems for monitoring, anomaly detection, root-cause analysis, and automated remediation. Solve hard problems in diagnostics, performance tuning, and capacity planning. Use safeguards such as rigorous evaluation, access controls, canary releases, and rollback so that automation stays safe.
5. Platform Architecture: Help design the next-generation cloud data platform for AI-native workloads, bringing storage, compute, retrieval, and model serving together. Own key modules or cross-module architecture, and make sound trade-offs across performance, reliability, openness, and cost.
6. Research to Production: Identify high-impact open problems across data infrastructure, AI infrastructure, and AI for infra. Apply ML, LLMs, agents, and reinforcement learning to real system problems, including RL reward modeling, policy optimization, and feedback learning. Train, fine-tune, and distill in-house models for data understanding, retrieval ranking, and system optimization. Take work from hypothesis to benchmarking to production, and share results through open source and top-tier conferences.
Who We Look For
1. Bachelor's degree or higher in Computer Science or a related field, with 5+ years of software engineering experience. You have strong fundamentals in data structures, algorithms, operating systems, and networking, and strong system design skills.
2.…