jobs in BeiBei

Kerja Sepenuh Masa, Artificial Intelligence Engineer di BeiBei - Maukerja

Artificial Intelligence Engineer

BeiBei

Singapore

Kongsi
Simpan

Lokasi Kerja

  • Singapore

Penerangan Kerja

Tanggungjawab

Responsibilities

  • Responsible for the design and development of intelligent Agent systems in the financial investment domain. Build AI application systems for real-world scenarios such as investment research, risk management, operations, and user services, with capabilities including task planning, tool calling, context understanding, and result validation.
  • Participate in the evolution of Agent architecture, including core capabilities such as Agent Runtime, Workflow, Tool Use, Memory, RAG, Multi-Agent collaboration, reflection and self-correction, evaluation, and observability. Improve the stability, accuracy, and controllability of Agents in complex, multi-step, and knowledge-intensive tasks.
  • Design and optimize multi-tool orchestration, task decomposition, multi-step reasoning workflows, context management, exception handling and fallback mechanisms, result verification, and human feedback mechanisms to drive Agents toward production-grade applications.
  • Participate in the development of AI platforms and infrastructure, including model gateways, Prompt / Skill management, knowledge retrieval, evaluation systems, log tracing, cost monitoring, and quality analysis, providing stable and efficient underlying infrastructure for Agent systems.
  • Collaborate closely with product, algorithm, backend, and data teams to integrate Agent capabilities into investment research data, trading workflows, risk management systems, and operational platforms, driving AI capabilities to directly impact business metrics.
  • Continuously follow the latest developments in LLM Agents, RAG, tool calling, code generation, multimodal interaction, performance optimization, observability, and engineering practices, and rapidly validate and implement emerging technologies in real-world business scenarios.

Requirements

  • Bachelor's degree or above in Computer Science, Artificial Intelligence, Software Engineering, or a related field. We value your passion for technology, understanding of business, and proven track record of solving complex problems through practical work.
  • Strong engineering and software development fundamentals. Proficient in at least one backend programming language among Python, Go, Java, Rust, and C++. Able to independently design and implement complex systems. Strong coding practices and solid knowledge of data structures and algorithms.
  • Familiar with the LLM and Agent technology ecosystem, with a deep understanding of LLM architectures such as Transformers and Attention mechanisms. Familiar with major open-source model ecosystems. Strong understanding of mainstream Agent paradigms including Tool Calling, Workflow, ReAct, Plan-and-Execute, Memory, and Multi-Agent systems. Hands-on experience designing and developing Agent systems, including autonomous planning, tool use / function calling, memory mechanisms, reflection, and self-correction, with proven experience integrating AI capabilities into real-world business workflows.
  • Excellent problem-solving, decomposition, and system design skills. Able to abstract complex and ambiguous business problems into scalable system architectures, reusable tool capabilities, and sustainable technical solutions.
  • Familiar with the complete RAG and knowledge retrieval pipeline, including document parsing, chunking, embeddings, retrieval, reranking, citation and provenance, and retrieval evaluation. Understands the applicable scenarios and trade-offs of different approaches and can make appropriate technology choices and optimizations based on business requirements.
  • Strong backend system design and engineering capabilities. Familiar with Python web frameworks such as FastAPI / Flask and understands API design, asynchronous tasks (asyncio), caching strategies, message queues, data storage, microservice architectures, access control, canary / gradual releases, monitoring, and alerting. Able to make appropriate technology choices based on business scenarios. Familiar with Docker / Kubernetes containerization and understands CI/CD practices.
  • Strong awareness of production system stability, with experience troubleshooting and resolving issues in production environments. Able to continuously optimize systems in terms of latency, throughput, cost, observability, fault recovery, and incident investigation.
  • Familiar with Prompt Engineering methodologies and capable of systematically designing and optimizing prompt strategies, with hands-on experience in evaluation and iterative improvement.
  • Strong curiosity and learning ability toward new technologies. Highly self-motivated, continuously follows industry best practices, and can rapidly apply new knowledge to practical projects.

Preferred Qualifications

  • Hands-on experience with AI Agents or related areas, having led or deeply participated in Agent design, tool orchestration, task planning, memory mechanisms, evaluation systems, or production deployment. Experience taking an Agent system from 0 to 1 and continuously operating and optimizing it in production is especially valuable.
  • Experience with AI platforms or AI infrastructure, such as model gateways, Prompt / Skill platforms, Agent Runtime, knowledge base platforms, retrieval systems, evaluation platforms, log tracing, or cost governance platforms.
  • Experience with in-house model fine-tuning, including SFT / RLHF / DPO, or pre-training. Familiarity with training data construction and evaluation processes, as well as distributed training frameworks.
  • Experience in performance analysis, observability, production incident troubleshooting, or system stability engineering. Able to identify system bottlenecks through logs, traces, monitoring metrics, load testing results, and other tools, and drive optimization.
  • Experience with information retrieval, vector search, search engines, recommendation systems, data platforms, or knowledge base systems. Understands the trade-offs between different storage, retrieval, and indexing solutions in terms of accuracy, latency, cost, and maintainability.
  • Experience with cloud-native or large-scale distributed systems. Understanding of Docker, Kubernetes, CI/CD, service governance, canary / gradual releases, elastic scaling, and other engineering practices.
  • Experience in code generation, automated software development, AI Coding, intelligent operations, automated data analysis, or related areas.
  • Live trading experience in U.S. equities, Hong Kong equities, or options, or a strong interest in company research, fundamental analysis, quantitative investing, or financial data systems. Able to understand financial application requirements from a user's perspective.
  • Familiarity with emerging financial product formats such as Event Contracts and prediction markets.
  • High-quality open-source contributions or sustained technical community engagement and influence, such as through blogs, research papers, or technical presentations.

About Our In-House Model Development

We not only leverage open-source and commercial foundation models, but are also actively exploring in-house model capabilities. In this role, you will have the opportunity to contribute deeply to the following areas:

  • Domain-Specific Model Training: Build high-quality training data for financial investment scenarios and participate in SFT and alignment training for domain-specific models.
  • Agent-Specific Model Optimization: Design specialized training approaches for core Agent capabilities such as tool calling, multi-step reasoning, and structured output.
  • Evaluation and Iteration: Build model evaluation pipelines and use automated evaluation to continuously drive model improvement.
  • Inference Optimization: Participate in model quantization, inference acceleration, and deployment optimization to ensure controlled latency and costs for Agent systems in production.


Peringatan Penting

Jangan pernah kongsikan maklumat bank atau kad kredit anda semasa memohon pekerjaan. Elakkan membuat sebarang pembayaran atau mengisi survey yang tidak berkaitan. Jika ada yang mencurigakan, sila laporkan iklan pekerjaan ini segera.

Lebih Lanjut