Quantitative Research | Trading Systems | AI-assisted Engineering
I build quantitative research and trading systems across A-share equities, domestic futures and crypto perpetuals. My work connects strategy hypotheses, data pipelines, backtest validation, execution monitoring and post-trade attribution using Python/QMT and event-driven frameworks.
I also support empirical financial research through event studies, DID, matched controls and robustness checks. Public repositories below are self-owned, sanitized showcases of research methods and engineering delivery.
Open to Quantitative Research, Quant Development, Research Engineering, Trading Systems and AI Workflow roles in Singapore, Hong Kong, the UAE and remote teams.
LinkedIn | Public repositories
| Project | What it demonstrates | Suggested review |
|---|---|---|
| quant-trading-system | A-share research infrastructure with backtesting, realtime market monitoring, industry heatmaps and paper-trading workflows. | Backtest logic, factor validation, data fallbacks, screenshots and regression tests. |
| super-pricing-system | Asset-pricing and macro research platform with CAPM/FF3/DCF experiments, macro-factor dashboards and evidence-quality diagnostics. | Pricing models, research APIs, workflow orchestration and frontend/backend integration. |
| index-inclusion-research | Empirical-finance toolkit for event studies, matched controls, permutation tests and clustered standard errors. | Research design, reproducible pipelines, robustness tables and limitations. |
| rayn-strategy | Offline crypto research scaffold with bounded-grid and momentum backtests, risk checks and public-data tooling. | Backtest engines, accounting assumptions, configuration profiles and regression tests. |
- A-share strategies: sector rotation, call-auction signals, CSI 500 enhancement and a multi-asset martingale-grid cash pool; model holdings, costs and execution constraints and compare against explicit controls.
- Domestic futures CTA: event-driven research with contract rolls, settlement, margin and close-today fees; cost-aware walk-forward validation, paper/backtest reconciliation and a CTP adapter.
- Crypto perpetuals: Freqtrade research with 1h signals and a 1m exit adapter; shared-wallet, fee, funding and fill/equity checks, plus Docker deployment, monitoring and recovery.
These systems include private research and company work; the public repos are separate showcases. Research backtests, paper execution and live performance are distinct forms of evidence.
I organize Codex, Claude Code, MCP and reusable skills/plugins around a concrete delivery loop:
- Break down the task and define trading rules, data assumptions and acceptance criteria.
- Use agents for implementation, debugging and code review.
- Check outputs with automated tests, regression comparisons and research/accounting validation.
- Review the evidence and document deployment or handover decisions.
I own the architecture, trading logic, research validity and final validation. AI-assisted implementation is supported by inspectable code and tests.
- Research: Python, SQL, pandas/NumPy, scikit-learn/statsmodels, causal backtesting, walk-forward validation, factor IC and event studies.
- Trading and engineering: QMT/xtquant, Freqtrade, FastAPI, React/TypeScript, PostgreSQL/TimescaleDB, Redis/Celery, Docker, GitHub Actions and Playwright.
- AI workflows: Codex, Claude Code, MCP, reusable skills/plugins, implementation review and regression validation.
- QuantTradingApp: Android stock-research demo with Kotlin, Room, local screening and review notes.
- etf-512400: local-first ETF research console with quotes, NAV checks, factor signals and source-health diagnostics.
- yieldwise: geospatial analytics with PostGIS, maps, candidate queues and local research memos.
Private/company code, credentials, account-connected trading systems and sensitive runtime data are not published here.