台北市中正區3年以上碩士以上
About Us:
This role is established to build and strengthen the company's core AI capabilities by leveraging advanced large language models (LLMs) and reinforcement learning (RL) technologies. The ultimate goal is to directly enhance our trading decision-making and market analysis capabilities.
You will collaborate with a cross-functional team of quantitative researchers and software engineers to develop the next generation of AI-driven financial products.
What You Will Learn:
- Practical applications of Agentic AI in the fintech domain
- Techniques for optimizing LLMs in terms of both accuracy and performance
Roles/ Responsibilities:
• Design end-to-end machine learning pipelines, including data collection, reprocessing, model training, and deployment.
• Develop and fine-tune large language models (e.g., LLaMA 3, DeepSeek R1) for financial market forecasting and sentiment analysis.
• Design and implement AI agent systems to enhance execution efficiency and accuracy of trading strategies through autonomous decision-making and interaction capabilities.
• Experiment with and implement cutting-edge LLM and reinforcement learning algorithms, utilizing techniques such as prompt engineering, RAG, MoE, LoRA, and TTS to continuously improve model accuracy and performance.
• Optimize model inference speed through techniques like quantization, distillation, and the use of efficient inference frameworks (e.g., TensorRT) to ensure real-time processing of large-scale financial data.
• Collaborate closely with quantitative researchers and data engineers to translate model outputs into actionable investment strategies.
• Continuously monitor model performance, conduct ongoing optimization, and troubleshoot issues as they arise.
Required Skills:
• Master's degree or above in Computer Science, Electrical Engineering, Data Science, or a related field is preferred.
• Minimum 3 years of hands-on experience in NLP or large language model (LLM) development.
• Proficient in Python, with experience in training models using PyTorch or TensorFlow.
• Familiar with the LLM ecosystem, including tools such as Cursor, Ollama, DeepSeek R1, Hugging Face, and LangChain.
• Experienced in large-scale data processing and modeling, with a solid understanding of distributed training and model optimization techniques.
• Strong problem-solving skills and a collaborative team spirit.
Preferred Qualifications:
• Experience in developing AI models for financial applications, such as market forecasting or sentiment analysis.
• Participation in AI-related competitions (e.g., Kaggle, AI Challenger) or contributions to well-known open-source projects.
• Publications in top-tier conferences such as NeurIPS, CVPR, or ICCV are a plus.
• Familiarity with cloud platforms (e.g., AWS, GCP) and model deployment processes is a plus.
Interview Process (Process sequencing may be adjusted as appropriate)
Resume selection -->Coding Assessment -> F2F or Google Meet Interview( Hiring team) --> HR Manager