Business Administration student focused on finance, M&A and investment research — building transparent analytical tools in Python and TypeScript.

I currently work in M&A / PMI at John Deere (Dealer Development) and previously worked at Prudentia M&A. This portfolio shows how I approach a finance question from first principles: define the decision, structure the data, make assumptions explicit, implement the analysis, test it and present an output that another analyst can review.

The repositories are research and educational tools, not investment advice. Their common theme is converting recurring finance workflows into transparent, reproducible analysis rather than black-box recommendations.

• Portfolio Intelligence Platform — Complete — A portfolio-research dashboard prototype covering allocation, risk, scenario analysis, watchlists and market intelligence. Earlier platform and advanced versions are archived development stages.
• Portfolio Optimizer — Constrained mean-variance, minimum-variance, risk-parity and Black-Litterman allocation with turnover-aware backtests and versioned reports.
• Portfolio Risk Dashboard — Historical VaR and Expected Shortfall, concentration, factor attribution, correlation regimes and stress analysis.
• Credit Risk Model — Probability-of-default modelling, calibration checks, explainability and expected-loss scenarios using explicitly documented synthetic data.
• Options & Volatility Lab — Option pricing, Greeks, implied volatility and Heston volatility-surface calibration.
• Walk-Forward Backtester — Out-of-sample strategy validation with execution frictions, benchmark comparison and regime analysis.
• Orderbook Microstructure Simulator — Market microstructure, execution quality and simulated order-book dynamics.
• Sentiment Market Signals — Text classification, event studies, lag tests, volatility analysis and cost-aware backtests, including negative findings.
• DACH Quant Optimizer — Quantitative research framework for German, Austrian and Swiss equity markets.

• Modern Automated DCF — Three-statement DCF scenario analysis with WACC, terminal value and sensitivity tables.
• Modern Automated LBO — Debt schedules, cash-flow waterfalls and MOIC / IRR scenario analysis.
• PE Deal Screener — Structured deal triage using an explicit mock-data path and versioned screening output.
• CIM Analyser — Rule-based extraction and structuring of information from mock CIMs, annual reports and investor presentations, with a human-validation gate.
• Financial Risk Screener — Earnings-quality, working-capital and disclosure-risk screening with a generated HTML review report.
• DCF Model Demo — Export-oriented DCF demonstration with versioned workbook outputs and a documented smoke check.

• Congressional Disclosure Monitor (Pelosi Stock Tracker) — Monitoring of delayed public congressional disclosures with market context. It is informational only and does not copy, score or recommend transactions.

Languages: Python, TypeScript and JavaScript

Finance & data: pandas, NumPy, SciPy, scikit-learn, cvxpy, Matplotlib and market-data APIs

Applications: React, tRPC, Express, Tailwind CSS and Vite Working method: Git, GitHub Actions, SQL, tests, versioned reports, explicit data scope and documented limitations

Private banking, wealth management and asset management require a broad view of markets, risk and client constraints. M&A and private equity require the same discipline from a transaction perspective: understand the business, challenge assumptions, identify risks and make the output reviewable. This is the capability I am developing through the portfolio: sound finance judgment supported by structured, maintainable technology.

GitHub: gilhermanns LinkedIn: gil-hermanns

Developed with support from Claude Code (Anthropic); the modelling choices, validation and documentation remain my responsibility.

The repositories are for research and educational purposes and do not constitute investment advice.