AI Agent Developer

CSOP Asset Management; Hong Kong; On-site; HKD 900,000 - 1,300,000; Full time

Job summary

CSOP Asset Management is seeking an accomplished AI Engineer to design and implement AI Agent systems that automate complex financial workflows, including trade support and data analysis. You will integrate LLM pipelines and orchestrate multi-step processes with internal data sources.

The role requires strong prompt engineering, model governance, and clear technical documentation. You'll work with traders, PMs, and compliance to deliver enterprise-grade AI capabilities in a fast-paced

Qualifications

• Bachelor's degree or above in a quantitative field (CS/Engineering/Math).
• Experience with LLMs, AI agents, and prompt engineering.
• High proficiency with modern AI development tools and environments.
• Strong English communication and clear technical documentation skills.
• Ability to decompose complex business problems into structured engineering tasks.
• Self-directed work style with ability to manage multiple initiatives.

Responsibilities

• Solution development for AI agent systems in financial workflows.
• LLM integration and orchestration across data sources and tools.
• Prompt engineering to ensure accuracy, consistency and compliance.
• Rapid prototyping and validation with end users and SMEs.
• Stakeholder engagement with traders, PMs, distribution, and compliance.
• Production handoff: documentation and runbooks for platform engineers.
• Technology evaluation and leadership briefings on AI capabilities.

Skills

AI agent frameworks

Large language models

Prompt engineering

English communication

Analytical thinking

Self-directed work style

Education

Bachelor's degree in CS/Engineering/Math

Tools

GitHub Copilot

Dify

Cursor

Claude

Job description

• Solution Development: Design, architect, and implement AI Agent systems that automate or augment complex financial workflows, including trade support, market research synthesis, client communications, and data analysis.
• LLM Integration & Orchestration: Build and maintain multi-step agent pipelines leveraging frameworks such as Dify, or equivalent orchestration platforms. Integrate with Bloomberg API, internal data infrastructure, and third-party data providers to deliver domain-specific AI capabilities.
• Prompt Engineering: Develop, evaluate, and optimize structured prompts and agent instructions to ensure accuracy, consistency, and compliance-readiness in LLM-generated outputs across financial contexts.
• Rapid Prototyping & Validation: Translate business requirements into functional prototypes on compressed timelines. Validate solution feasibility through iterative testing with end users and subject matter experts.
• Stakeholder Engagement: Partner with traders, portfolio managers, distribution professionals, and compliance officers to identify automation opportunities, gather requirements, and ensure delivered solutions meet institutional quality standards.
• Production Transition: Prepare comprehensive technical documentation, architecture diagrams, and operational runbooks to support handoff of validated solutions to the platform engineering team for production deployment and ongoing maintenance.
• Technology Evaluation: Monitor and assess emerging AI capabilities, tooling, and best practices. Provide structured recommendations to senior leadership on technology adoption and strategic direction.

Qualifications

• Bachelor's degree or above in Computer Science, Engineering, Mathematics, or a related quantitative discipline.
• Demonstrated experience with large language models, AI agent frameworks (AutoGen, CrewAI, Dify, or comparable), and prompt engineering methodologies.
• High proficiency with AI-assisted development environments including GitHub Copilot, Cursor, Claude, and GPT-based tooling.
• Excellent English communication skills — both written and verbal — with the ability to produce clear technical documentation and precise model instructions.
• Strong analytical and problem-solving capabilities; ability to decompose complex, ambiguous business problems into structured, deliverable engineering tasks.
• Self-directed work style with the ability to manage multiple concurrent initiatives and deliver under time constraints.