LLM/AI Ops Development Engineer Graduate

About the Role

The Network Observation team is building a world-leading hyperscale data center network infrastructure to support hundreds of millions of users' real-time access and massive data volume growth. We believe the next generation of network operations will be powered by artificial intelligence technologies, particularly Large Language Models (LLMs). We are seeking a passionate development engineer with deep networking expertise and innovative AIOps capabilities to help define and build "autonomous" data center networks. This role will transform network operations into a proactive, data-driven intelligent ecosystem with predictive and self-healing capabilities. As a graduate, you will have opportunities to pursue bold ideas, tackle complex challenges, and unlock limitless growth. Successful candidates must commit to an onboarding date by the end of the year.

Key Responsibilities

• Design and implement a closed-loop AIOps for Network platform.
• Build a Panoramic Network Observability Platform: Develop a streaming telemetry data pipeline for physical and virtual networks, integrating multi-source data from gNMI, Netconf, IPFIX/NetFlow, and SNMP for a high-quality, real-time data foundation.
• Develop an Intelligent Diagnostics and Root Cause Analysis System: Apply machine learning and deep learning algorithms for anomaly detection, correlation analysis, and intelligent noise reduction on network metrics, logs, and events. Swiftly pinpoint root causes of failures across the entire stack.
• Explore Innovative Applications of LLMs and Agents:
• Intelligent Operations Assistant: Build a conversational chatbot powered by Retrieval-Augmented Generation (RAG) for natural language query understanding, knowledge base querying, and providing troubleshooting guidance and network status reports.
• Automated Remediation and Smart Runbooks: Train operational Agents to safely and controllably invoke network change tools and APIs, enabling them to autonomously generate, recommend, or execute remediation plans and emergency runbooks.
• Establish Capacity and Risk Prediction Capabilities: Forecast network capacity bottlenecks, high-risk links, and "sub-healthy" devices based on historical data and business growth models for proactive scaling and preventative maintenance.
• Forge a Rock-Solid Engineering System: Adhere to engineering best practices to design and develop a highly available and scalable AIOps platform, guaranteeing stability and performance from data collection and model training to online inference and automated actions.

Requirements & Tech Stack

• Deep understanding of data center network architectures (e.g., Spine-Leaf Fabric).
• Proficiency in key protocols such as EVPN/VXLAN and BGP/OSPF.
• In-depth knowledge of the Linux network stack.
• Mastery of Golang or Python with outstanding coding and system design abilities.
• Familiarity with modern software development workflows, including microservices, containerization (Docker/Kubernetes), and CI/CD.
• Practical experience in one or more of the following areas is highly desirable:
• Big Data Processing: Familiarity with Kafka, Flink, ClickHouse/TSDB, and experience building real-time data pipelines and analytics systems.
• Observability Technologies: Experience with Prometheus/OpenTelemetry, graph databases (e.g., Neo4j), and developing alert and event platforms.

Qualifications

• Individuals completing or recently completed a Bachelor's or Master's degree in Computer Science or a related discipline.
• Experience in operating or developing for hyperscale (100,000+ servers) data center networks.
• Proven experience leading or making significant contributions to an LLM/Agent-based intelligent operations project with measurable business impact.
• Active contributions to open-source communities such as SONiC, P4/PINS, eBPF, Prometheus, or OpenTelemetry.
• In-depth research or practical experience in high-performance networking (RDMA/RoCE), SmartNICs (NIC Offload), or DPDK/eBPF.
• Experience building network configuration and control systems (e.g., based on SONiC, gNMI, Netconf).