Nassim Berrada

Conducts independent research on learning agents, reasoning, continual learning, and representations underlying agent behavior. Reviews current research, formulates hypotheses, designs experiments, implements research prototypes, and analyzes experimental results.

Designed and built an LLM evaluation platform for systematic benchmarking of API endpoints against custom benchmark suites. Built an LLM observability system for capturing and analyzing complex model inputs and outputs, including execution traces, structured data, diffs, and visual analytics.

Extended the Ivy machine learning framework and transpiler with PyTorch, JAX, and NumPy functionality across different numerical ecosystems. Developed a machine learning model for automatic generation of Ivy code, covering model design, training, evaluation, experimentation, and integration into the framework.

Built a multi-agent market simulation environment with heterogeneous learning agents to investigate systematic trading strategies and emergent market behavior. Developed machine learning pipelines for sensor data, covering data processing, feature engineering, model training, and evaluation.

Developed a real estate investment modeling dashboard for the alternative investments management division. Automated financial reporting and consolidation workflows to improve the efficiency and consistency of investment analysis.

Conducted market research on REITs and performed due diligence on prospective real estate investments.

Assessed the impact of changes to a pension division's analytical model on management-reported performance metrics.

Completed the Chartered Financial Analyst (CFA) Program from 2019 to 2022.

Pursued a Master's in Finance at ISCAE Business School from 2015 to 2018.