AI Engineer

Job Description

**Location**: Chengdu, China

---

**Company Overview:**

MetaApp, founded in 2017, is a rapidly growing quasi-unicorn technology company dedicated to building a leading interactive content creation and consumption platform, with a strong focus on the metaverse. Our mission is to "Expand Humanity’s Life Experiences," enabling users to explore countless virtual lives through innovative technology and products.

Our portfolio includes China's leading mobile game aggregation platform "233乐园" (233乐园), and the groundbreaking user-generated content (UGC) virtual world platform "MetaWorld" (along with its companion editor, "口袋方舟" - Pocket Ark). Leveraging our proprietary sandbox technology, 233乐园 has served over 200 million users, facilitating over a billion game distributions, and achieving industry-leading user activity and stickiness. MetaWorld aims to build an "all-ages Roblox-like" platform, empowering users and developers to create games, and deeply integrating AI technologies (such as MetaGPT) to enhance creative efficiency.

MetaApp has successfully secured multiple rounds of top-tier venture capital funding from investors including SIG and DST Global, with a valuation approaching $1 billion. We are financially robust, having achieved significant revenue and operational profitability. Our team is passionate, pragmatic, and high-performing, guided by our core values: "Data-Driven," "Direct Communication," and "Closed-Loop Accountability." Join us to collaborate with industry elites and shape the future of internet entertainment within a vast and exciting market.

---

**Role Overview:**

We are seeking an exceptional AI Engineer to join our Core Services team. In this pivotal role, you will be instrumental in tackling performance bottlenecks within our machine learning and deep learning systems. You will focus on the development and optimization of high-performance AI model inference service frameworks and engines, significantly improving real-time inference latency and offline batch processing throughput to meet our products' escalating performance demands. Your contributions will directly influence the successful delivery and impact of critical projects.

---

**Job Scope / What You'll Do:**

* **Design & Develop High-Performance AI Inference Service Frameworks:** Participate in the design and implementation of high-performance inference service architectures for large-scale AI model deployment. This includes developing and optimizing modules for model loading, unloading, version management, dynamic scaling, fault tolerance, high availability, and security.
* **Optimize Large Model Inference Engine Performance & Efficiency:** Research and apply heterogeneous machine inference engines like TensorRT, ONNX Runtime, and OpenVINO. Conduct operator optimization, graph optimization, and parallel strategy optimization for large models (e.g., GPT, Transformer-based models), explore and implement model compression techniques (e.g., quantization, distillation), and perform performance tuning across multi-hardware platforms (CPU/GPU).
* **Evaluate, Integrate & Deploy AI Infrastructure Components:** Research and evaluate industry-leading AI infrastructure components (e.g., feature platforms, model management systems, A/B testing platforms). Be responsible for integrating selected components into existing platforms and driving their adoption in practical business scenarios.
* **Solve Cutting-Edge ML System Challenges:** Continuously monitor the latest advancements and academic research in Machine Learning Systems (ML-System). Analyze complex ML system problems encountered in real-world business scenarios and propose innovative solutions. Participate in internal tech seminars to share research findings and practical experience, particularly in areas like ultra-large-scale model deployment, heterogeneous computing optimization, and multimodal model inference.
* **Maintain & Iterate on Existing ML/DL Platforms:** Oversee the daily monitoring, troubleshooting, and performance diagnosis of existing machine learning/deep learning platforms. Implement functional iterations and upgrades based on business requirements and technical planning, and continuously optimize platform code structure for quality and maintainability.
* **Build & Maintain ML Platform Data Pipelines & Monitoring Systems:** Design and implement data pipelines for model training and inference, as well as robust monitoring and alerting systems.
* **Collaborate Closely with Algorithm Teams:** Regularly communicate with business algorithm teams to deeply understand model inference performance requirements in various business scenarios, and assist in resolving technical issues encountered during model deployment and integration.

---

**Requirements / Who You Are:**

**Minimum/Core Requirements:**

* 3+ years of experience in AI/Machine Learning systems, platform, or inference engine development.
* Proven practical experience in high-performance computing or large-scale distributed systems, familiar with high-concurrency, low-latency, and distributed training/inference scenarios.
* Exceptional C++ (C++11/14/17) programming skills, with the ability to write performance-sensitive code and perform deep performance tuning.
* Proficiency in at least one scripting language among Python, Shell, or Lua.
* Experience in large model (e.g., Large Language Models, Transformer-based models) inference optimization, with successful outcomes in improving throughput or reducing latency.
* Demonstrated success in designing, developing, and deploying production-grade, high-performance machine learning inference service frameworks, including end-to-end project experience.
* In-depth research and innovative solutions to complex, cutting-edge ML system challenges, potentially evidenced by publications or open-source contributions.
* Practical experience leading or significantly contributing to the maintenance and iteration of machine learning platforms, including daily operations, troubleshooting, and feature iterations.
* Solid understanding of operating system principles, such as processes/threads, memory management, file systems, and networking.
* Familiarity with CUDA/NPU programming for heterogeneous computing acceleration and optimization.
* Understanding of model compression principles (e.g., distillation, quantization) with practical application experience.
* Excellent learning agility and enthusiasm for new technologies, able to quickly adapt to new tech stacks and challenges.
* Strong independent problem-solving abilities and outstanding system architecture design skills.
* Proactive communication and cross-functional collaboration skills, especially with business algorithm teams, to align technical solutions with business goals.
* Deep understanding of the machine learning/deep learning domain, including foundational theories and common algorithms.
* Profound understanding of the architecture and characteristics of large models (e.g., GPT, Transformer-based models).
* Strong sense of responsibility and ownership, striving for technical excellence and high-quality deliverables.
* Ability to thrive in a fast-paced technical environment and eager to embrace challenges.
* Bachelor's degree or higher in Computer Science, Artificial Intelligence, Mathematics, Physics, or a related engineering field.
* Ability to work fully in-office and be based in Chengdu.

**Preferred/Bonus Requirements:**

* Experience in the R&D of specific model inference engines (e.g., TensorRT, ONNX Runtime, OpenVINO).
* Experience in selecting and driving the end-to-end implementation of AI infrastructure components (e.g., feature platforms, model management systems) from scratch.
* Proficiency with common deep learning models beyond large models, including those used in Computer Vision (CV) or Natural Language Processing (NLP).
* Familiarity with data pipeline and monitoring system setup, such as Kafka, Prometheus, Grafana.
* Familiarity with the gaming or metaverse industry.
* Active participation in open-source communities, with contributions to popular open-source projects.

---

**What We Offer:**

* **Vast Industry Opportunities & Impact:** Be at the forefront of the mobile gaming and metaverse industries, helping to build the next generation of interactive entertainment. Your work will directly impact hundreds of millions of users.
* **Top-Tier Team & Personal Growth:** Collaborate with industry-leading technical and product talents. Benefit from extensive learning and growth opportunities, including comprehensive onboarding programs, mentorship, internal tech sharing sessions, and book clubs.
* **Innovative Culture & Career Progression:** Thrive in a data-driven, flat, and efficient culture that encourages bold innovation and rapid iteration. As a fast-growing company, we offer ample opportunities for career advancement and taking on greater responsibilities.
* **Competitive Compensation & Benefits:** Enjoy an above-industry-average salary, semi-annual performance reviews with salary adjustments, and generous year-end bonuses. Comprehensive benefits include social insurance (五险一金), supplementary commercial medical insurance, daily meal allowances, communication subsidies, transportation subsidies, and housing subsidies in cities like Beijing.
* **Work-Life Balance:** We operate on a 5-day work week with weekends off and flexible working hours. We provide national statutory holidays, paid annual leave (including additional welfare annual leave), sick leave, maternity leave, paternity leave, and special occasion benefits (e.g., birthday/wedding gifts), promoting efficient work and a healthy life.
* **Comfortable Office Environment:** Work in a spacious and comfortable office equipped with high-performance computers, dual monitors, and other necessary tools. Enjoy complimentary coffee, tea, snacks, and leisure areas with table football, board games, and console games for relaxation and team bonding.

If you are passionate about making a significant impact in a rapidly evolving industry and eager to build groundbreaking products with a top-tier team, we encourage you to apply now and help MetaApp "Expand Humanity’s Life Experiences"!