Join an early-stage, high-velocity team building AI agents that automate end-to-end compliance investigations for banks.
The product is already live, and you'll work directly on the core platform-building and evaluating agent systems that power real compliance investigations.
This is a hands-on engineering role with real ownership, working on production systems from day one.
6 Months Internship.
Full time offer based on Performance.
Requirements:
- Build and improve agent harness for compliance investigation use cases
- Develop and evaluate LLM-based extraction, classification, and reasoning components
- Contribute to synthetic data generation for agent testing and evaluation
- Write evaluation harnesses to measure agent accuracy and reasoning quality
- Document agent behaviour, evaluation results, and design decisions
- Track latest agentic AI research and bring relevant techniques into production (context management, long-term memory, agent learning)
- Train and fine-tune machine learning models from scratch for compliance-specific use cases
- Robust foundation in Python; comfortable with ML libraries
- Exposure to or strong curiosity about LLMs and agentic frameworks
- Independent – able to own problems, not just sub-tasks, in a small team
- Good written communication; able to document and present your work clearly
- Prior coursework or projects in ML, NLP, or distributed systems is a plus
- No prior fintech or compliance experience required