c:I[5516,["591","static/chunks/591-af5be930573c970b.js","296","static/chunks/app/station-details/%5BstationID%5D/page-05d8540f5449cd68.js"],"default"] d:T417,

Description:

This project focuses on optimizing metrics computed on portfolios and banking books, such as capital

consumption, risk-weighted assets, or concentration limits. The student will assist in building tools to

optimize metric outcomes under various constraints and input sensitivities. This includes cleaning and

transforming financial data from Excel, structuring optimization problems, and prototyping algorithms

that improve metric efficiency.

Required Skills & Traits:

Strong programming skills in Python

Familiarity with Excel file handling (e.g., openpyxl, pandas)

Understanding of mathematical functions and basic optimization

Knowledge of portfolio or banking book structure (preferred but not mandatory)

Expected Outcomes:

Tools to automate metric computation and track optimization progress

Prototype optimization strategies applied to real banking book data

Framework to experiment with different constraints and metric targets

e:T463,

Description:

This project aims to develop machine learning models that approximate complex pricing routines for

financial instruments. These models will serve as fast surrogates to significantly accelerate risk

computations performed across large market scenario sets. The student will work with historical pricing data, test various regression or neural network approaches, and evaluate accuracy vs. speed trade-offs.

Required Skills & Traits:

Strong Python programming and numerical computation skills

Familiarity with NumPy, SciPy, and scikit-learn or PyTorch

Understanding of basic derivatives pricing and risk metrics (helpful but not essential)

Analytical mindset and comfort with model evaluation techniques

Expected Outcomes:

Cognext Analytics Private Limited - Analytics / Optimization

Station Overview

Project 1

Title: Portfolio-Level Metric Optimization for Banking Books

Project 2

Title: Machine Learning Surrogates for Pricing Financial Instruments