Selected public-facing project themes from my quantitative research and research-engineering work. Descriptions are intentionally generalized to avoid disclosing proprietary data, research logic, system architecture, or employer-specific implementation details.
Crypto Forecasting Research Framework
A modular workflow for turning raw market data into model-ready research experiments.
- Built reusable components for data preparation, feature evaluation, model training, and diagnostic reporting
- Added checks around missing data, asset histories, and experiment assumptions to improve research reliability
- Structured the workflow so model changes and validation results can be reviewed consistently
Backtesting and Experiment Validation Toolkit
A set of research tools for evaluating hypotheses under explicit assumptions and repeatable test conditions.
- Designed validation checks for timestamp alignment, universe construction, and data availability
- Improved configuration patterns so experiments could be compared without hidden state or manual changes
- Expanded regression tests for edge cases that commonly affect quantitative research results
Event-driven Market Research System
Infrastructure for collecting, replaying, and analyzing event-driven market data in a controlled research setting.
- Implemented components for data recording, replay, and offline analysis
- Separated research logic from production-facing code paths to make experiments easier to audit
- Used performance diagnostics to identify bottlenecks and improve iteration speed
Binary-payoff Pricing and Risk Analysis
A research framework for thinking about bounded-payoff contracts, implied probabilities, and risk-aware valuation.
- Applied option-pricing and probability-based reasoning to binary payoff structures
- Compared model-implied values with market prices under transparent assumptions
- Built monitoring views for valuation, exposure, and scenario analysis without exposing decision rules
AI-assisted Research Engineering Workflow
A practical workflow for using language models to support quantitative research engineering while keeping human review and tests central.
- Used structured prompts to clarify assumptions, edge cases, and expected behavior before implementation
- Paired code generation with iterative review, failure-mode analysis, and targeted tests
- Applied the workflow to research utilities, validation logic, and maintainability improvements
Blockchain Incentive and Data-market Analysis
Game-theoretic research on incentives in blockchain systems and data-property-right allocation.
- Reviewed economic literature on blockchain incentives and data markets
- Built formal models to reason about participant behavior under asymmetric information
- Evaluated welfare implications across alternative mechanism assumptions