AHAD HAMIRANI

Hello! I'm Ahad

AI Engineer & Data Scientist with 2.5+ years of experience in predictive analytics, credit risk, and fraud detection using Python, PyTorch, and Scikit-learn. Engineered a national-level bureau credit score utilized across Indian lending institutions and architected hierarchical models to predict fraud. Leverages MSCS and an MBA to optimize financial decision-making through high-throughput model deployment and scalable, end-to-end ML solutions that align deep learning with strategic business growth.

Ahad Hamirani

Georgia Institute of Technology | Ex-Experian | NMIMS'22

+91 9820186663

Phone:

Email:

ahadhamirani62@gmail.com

Date of Birth:

August 4th, 1999

EXPERIENCE

JUN 2022 - Feb 2025

Data Modeler II

EXPERIAN CREDIT SERVICES INDIA

Engineered an updated bureau credit score, currently utilized across all Indian credit lending institutions to assess credit risk.
Architected a Mule Risk Indicator for bank’s liability accounts using a hierarchical modeling approach, significantly reducing the bank’s liability fraud and mis-selling instances.
Synthesized industry credit insights and trend analyses to guide financial lending decisions and strategic market positions.
Developed end-to-end acquisition scorecards and propensity models for diverse clients, engineered predictive features from bureau data to automate customer journey and optimize the risk strategy.
Oversaw model management and client integration for “Hunter” fraud detection score (Pro and Prime), ensuring seamless deployment and consumption for external stakeholders.
Named finalist in IRDAI’s “Bima Manthan” Hackathon for developing an innovative transparency solution aimed at curtailing mis-selling within the insurance sector
Modules/Technology used: Python, Scikit-Learn, Bitbucket

FEB 2021- NOV 2021

JOCATA FINANCIAL & ADVISORY

Machine Learning Engineer

Worked in the Data Analysis Department as a Machine Learning engineer on the Loan operating system using NLP, Deep Learning, NN, and transfer learning to determine the amount of loan a person is eligible for and their financial stability, using their passbooks and analyzing their transaction pattern providing a detailed report.
Designed and developed a tool for extracting graphical features on transactional data, used in conjunction with features extracted from a rule-based engine to detect money laundering, analyzing the flow of money in the network to detect frauds.
Modules/Technology used: Neo4j (Graphical Database), Python, Tensorflow, PyTorch

APR 2020 - JUL 2020

Machine Learning Engineer

JOCATA FINANCIAL & ADVISORY

Worked in the Data Analysis Department as a Machine learning engineer and designed an Adverse News Media Analysis tool using Natural Language Processing (NLP), to analyze an entity’s market sentiment and the trend of negativity based on the news and articles published for it. Used to identify the general sentiment of the entity and find out its financial stability.
Modules/Technology used: TensorFlow, Python, Django, REST API

MAY 2018- JUN 2018

Software Developer

CRISIL Limited

Worked as a developer in the Risk solution technology on an Early Warning System to detect default entities in their loan repayment using a rule-based engine, with a manual override based on the hierarchy. used by banks such as ICICI and SBI.
Modules/Technology used: SQL, Java, spring framework

EDUCATION

2024-2026

MBA in Technology Management

Georgia Institute of Technology

GPA 4.0/4.0

2017-2022

MBA in Technology Management

NMIMS Deemed-to-be-University

GPA 3.53/4.0

2017-2022

Btech in Computer Engineering

NMIMS Deemed-to-be-University

GPA 3.53/4.0

INTERESTS

RESEARCH PROJECTS

RESEARCH PROJECT

Mental Health Digital Intruder Project

Working in an International Mental health digital intruder project as a coordinator and a member, the project aims to detect and help people suffering from mental illnesses based on the data collected from the three modules in the app (Game, Music, and Social media)

Partially Occluded Object
Detection and Classification in
Video Sequences

Management Analytics Software

Using various image processing techniques and deep learning models like YOLO and Mask RCNN to recognize and detect partially occluded objects in a video feed, also recognizing occlusion using various metrics.

Developed a web-based analytical software to reduce the redundant tasks performed daily in corporations, four modules were incorporated, finance, marketing, operations, and hr. Each module focused on tasks related to that branch, e.g. product sentiment analysis, forex prediction, man-hour calculation, skill chart visualization

SKILLS