Gero
Bone-
Winkel

Berlin — TU Berlin / EPFL

Work
2025
Technical Lead & Core Developer
VisualVest × TU Berlin
  • Graph Transformer pipeline extracting investment signals from earnings calls via OpenAI API (Python, FastAPI, Neo4j)
  • RAG architecture for structured document retrieval and signal aggregation
2023–24
First Author & Researcher
Technische Universität Berlin
  • Explainable ML survival analysis for P2P credit risk — published in Digital Finance
  • +15.63% IRR vs. benchmark; XGBoost + SHAP + Cox model
2024
Capital Markets Analyst
Aventos GmbH
  • Financial analysis of Berlin’s largest logistics development project
2022–23
Software Developer
Corussoft GmbH
  • AWS backend (Kotlin, TypeScript, SQL) for B2B enterprise applications
2020–22
Teaching Assistant
TU Berlin — Systems Programming & Computer Organization
Research
Digital Finance · 2024

Improving credit risk assessment in P2P lending with explainable machine learning survival analysis

XGBoost · SHAP · Cox Model · Survival Analysis · P2P Lending

↗
Vol. 6, pp. 501–542 doi:10.1007/s42521-024-00114-3
Education
2024 –
M.Sc. Industrial Engineering — Computer Science & ML
Technische Universität Berlin · GPA 1.0 (4.0 US)
2025–26
Exchange — Financial Engineering
École Polytechnique Fédérale de Lausanne (EPFL)
2018–23
B.Sc. Industrial Engineering — Computer Science
Technische Universität Berlin · GPA 1.3 · Top 3% of program
2021
Exchange — Technology Management Program
UC Santa Barbara
Skills
Languages
Python · Kotlin · C++ · TypeScript · SQL
ML & Data
XGBoost · SHAP · Optuna · NumPy · Pandas · Survival Analysis
Infrastructure
Docker · FastAPI · Neo4j · AWS · Git
Research
Explainable AI · Credit Risk Modelling · Graph Neural Networks · RAG