AI Engineer - Capital Markets Trading Intelligence
AI Engineer - Capital Markets Trading Intelligence
viseo asiaFresher
- Posted 12 hours ago
- Be among the first 10 applicants
Job Description
We're building an AI-powered tool that automatically delivers real-time trading information and insights, starting with rollout to institutional clients before expanding to private banking and wealth management. This is a high-visibility, cutting-edge role at the intersection of capital markets and applied AI.
Responsibilities
- Design and build an AI system that surfaces trading information automatically, tailored to the needs of institutional, private banking, and wealth management users
- Implement MCP (Model Context Protocol) integrations to connect the system to live trading data feeds
- Build RAG (Retrieval-Augmented Generation) pipelines against internal banking systems to ground outputs in proprietary data
- Perform post-training (fine-tuning / adaptation) on an existing LLM to specialize it for capital markets use cases
- Design, build, and run benchmarks to evaluate model performance against real-world trading and banking use cases
- Own the rollout strategy across client segments, iterating based on feedback from each phase (institutional → private banking → wealth management)
Profile
- Minimum Bachelors degree in IT/Computer Science or other relevant discipline
- Strong capital markets domain knowledge: you need to understand trading concepts, market data, and institutional/private banking workflows
- Hands-on experience post-training or fine-tuning LLMs
- Experience building RAG systems, ideally against enterprise/internal data sources
- Familiarity with MCP or similar protocols for connecting LLMs to live external data
- Experience designing evaluation frameworks and benchmarks for LLM-based systems
- Comfort working in a fast-moving, ambiguous environment building something genuinely novel
More Info
Key Skills
MCP Model Context Protocol
RAG Retrieval-Augmented Generation pipelines
capital markets domain knowledge
