As a Data Scientist with experience in banking and procurement systems, I specialize in building and delivering high-impact machine learning models that solve real-world business problems. My services include:
Credit Risk Modeling: I have developed and deployed models such as Application Scoring, PD estimation, and Limit Management systems. My models have achieved up to 78 Gini, ensuring strong predictive power and regulatory robustness.
Procurement Optimization & Recommendation Systems: I’ve built supplier recommendation systems using LightFM and Factorization Machines, enabling buyers to identify the best-matching suppliers based on product codes (UNSPSC), bidding history, and financial attributes.
NLP and Sentiment Analysis: I’ve developed an audio sentiment classification model for customer service calls using signal processing and ML, achieving measurable improvements in quality monitoring.
Data Analytics & Feature Engineering: I provide deep expertise in Python, SQL, Pandas, SHAP, optbinning, and Dataiku to clean, transform, and analyze complex data structures — whether financial, behavioral, or procurement-based.
Model Monitoring and Automation: I set up automated pipelines for performance tracking (e.g., PSI, Gini, drift detection), and integrate business logic and overlays where required.
What You Can Expect:
Clear and frequent communication
Clean, well-documented code
Models built with business impact in mind — not just academic accuracy
Support with deployment, interpretation, and visualization
Whether you're a fintech startup, a procurement firm, or a data-driven company looking to scale your AI capabilities — I bring both the technical skill and the business understanding to deliver value fast.