CV & Certificates

📧 abhirup.ghosh.184098@gmail.com

Senior data science leader with 6+ years driving high-impact ML strategies in fintech, mobility and astrophysics — shaping data science roadmaps, scaling AI-driven products, and delivering measurable business impact at organisational and multi-country levels. Recognised contributor to Nobel Prize-winning research, now leading advanced anti-financial-crime initiatives at Germany’s most valuable tech startup.

Current role

Senior Data Scientist, Anti-Financial Crime, Trade Republic — Berlin, Germany

Strategic leadership: led the AFC data science strategy for money muling, customer risk assessment, child savings accounts, risky KYC detection, and market launches across 5+ countries, driving compliant geo-expansion.
Technical innovation: built and deployed the first end-to-end GNN-based AML/CFT models to combat money muling and risky KYC/onboarding — a 60–80% improvement in detection rate, up to an 80% reduction in manual reviews, and a 21–40% reduction in laundered amounts — and the first machine-learning-based cards transaction monitoring system (+55% precision, +32% recall, −35% laundered amounts).
Governance: established BaFin- and ECB-aligned model governance and customer-risk-assessment standards.
Mentorship: grew data science capability through mentorship and senior-level hiring influence.

Previous roles

Senior Analytics Manager — Urban Mobility, Emmy / GoToGlobal — Berlin, Germany (05/2022 – 03/2024)

Strategy & BI: owned BI, analytics, and ML strategy for German operations — aligned initiatives with C-level and mentored 3+ analysts across 2 countries.
M&A integration: directed post-Felyx-acquisition data integration (+150% assets/users), shaping fleet, pricing, retention strategy, and subscription-model launch.
Revenue optimisation: architected AWS-based demand forecasting (sub-1% error), boosting fleet utilisation and revenue.
Cost management: consolidated 10+ data sources, cutting infrastructure costs by >50%.
Deployed CNN-based fraud detection (−20% fraud) and CV/OCR license verification (≈€100K annual savings).

Data Scientist / Researcher — Astrophysics, Max Planck Institute for Gravitational Physics (Albert Einstein Institute) — Potsdam, Germany (10/2018 – 04/2022)

Big data pipelines: led a 10+ member team to develop ML pipelines for TB-scale time-series data, contributing to ~90 gravitational-wave discoveries.
Production engineering: engineered workflows in the LIGO Algorithms Library, now core to global gravitational-wave data processing.
Extraordinary discoveries: architected anomaly-detection frameworks for “extraordinary” events, influencing LIGO-Virgo priorities — authored 100+ papers (60K+ citations, h-index 73), including the 2017 Nobel-winning landmark discovery paper of gravitational waves.
Mentorship: mentored 5+ PhD candidates, served on editorial/review boards, and set data-quality standards.

Key achievements

AML/CFT transformation: led the GNN roadmap driving a 60–80% detection-rate improvement, up to −80% manual reviews, and −21–40% laundered amounts at Trade Republic; led all account monitoring for 10M+ customers.
Analytics & BI scale-up: consolidated 10+ data sources, cut infrastructure costs by >50%, and enabled real-time insights at Emmy Sharing.
Business-impact modeling: delivered <1%-error demand forecasting and deep-learning fraud-detection (CNN, CV/OCR) systems, generating over €100K+ in annual savings.
Academic breakthroughs: co-authored the 2017 Nobel Prize-winning LIGO discovery paper; developed TB-scale anomaly-detection pipelines now core to global gravitational-wave operations.

Education

PhD, Astrophysics (gravitational-wave physics) — International Centre for Theoretical Sciences (ICTS), Bangalore, India (advisor: Prof. Parameswaran Ajith). Built the first-ever anomaly-detection pipeline with regression models of black holes using TB-sized audio-frequency time-series data from the LIGO-Virgo gravitational-wave experiment — co-authored the paper that won the 2017 Nobel Prize in Physics.
M.Sc. Physics — IIT Roorkee (2010–12, Rank 2)
B.Sc. Physics — Presidency College, Kolkata

Selected awards: GWIC-Bracchini Thesis Prize, Honourable Mention (2019) · Ramakrishna Cowsik Medal, TIFR India — best paper by a researcher under 35 (2017) · Special Breakthrough Prize in Fundamental Physics (2016)

Technical skills

Leadership & strategy: cross-functional leadership, product & AI strategy, strategic roadmapping, stakeholder management, mentorship & team development
Business & domains: AML/CFT compliance, fraud detection & risk scoring, mobility & revenue optimisation, demand forecasting & pricing, investment & stock-market analytics
Machine learning: statistical modeling, causal inference, Bayesian methods, forecasting, anomaly detection, graph algorithms (GNNs)
Platforms & MLOps: Python, SQL, AWS, GCP, MLflow, Airflow, Metaflow, Docker, Kubernetes, Claude Code, Cursor, Copilot
Data visualisation: Excel, Looker, Metabase, Power BI

Languages: English (C1), German (B2), Bengali (C2), Hindi (C1)