Daniel Milner Email & Phone Number

San Francisco Bay Area

Collaborated with the research scientist at one of the biggest tech company in San Francisco to optimize the referral program of their cloud services using Machine Learning – created a classifier to calculate the likelihood of people joining the cloud service using the referrals.

• Developed and designed a smaller version of Street fighter game using Pygame in Python, in which the game is controlled by the user actions in the webcam. User actions are classified into – run, kick, punch - by the.
• Implemented Nvidia’s End to End Learning for Self-Driving Cars CNN architecture on Pytorch to solve a challenge released by Comma.ai (an autonomous driving startup) to predict speed of a car from a video; achieved.
• Collaborated with Radiology Department of Northwestern University and implemented a 3D CNN architecture to predict likelihood of Parkinson’s Disease from brain MRIs.

Gurgaon, India

• Responsible for developing Machine Learning models and implementing Data Science techniques for travel management unit of Amex.
• Led a team of three to develop an Gradient Boosting Model to predict the attrition risk associated with 6000 clients, improving client retention by 6 times and thus saving $10 MN+ in revenue.
• Designed a clustering study (k-means and hierarchical) to target high-worth prospects of Amex GBT, with a potential increase of $50 MN+ in Revenue.
• Developed more robust Random Forest Model to replace traditional standard regression model to predict the volume won by American Express, increasing the prediction accuracy by 10%.
• Optimised a binary classifier using another Ensemble Technique (AdaBoost ) for the classification of American Express finance vendors.
• Collaborated with the US Director of the Performance and Insights team to pilot a critical project to realize the difference between actual transactions and potential transactions of 100+ clients with value of $1MN+.

Bangalore

• Credit Risk Analysis for one of the major European banksAnalyst, Impairment & Forecasting Swindon, UK
• Enhanced and Validated Risk management models - PD, LGD, EAD to increase the risk mitigation power of the bank - classification and dimensionality reduction.
• Identified lapses in Unidentified Impairment (Losses) approach to calculate impairment for the bank; redefined the methodology and presented the final recommendations to the Chief Risk Officer.
• Revised the segments of risk classification for High Risk Account Management team.Analyst, Credit Risk Analytics Bengaluru, India
• Recommended revising the write-off policy from 18 months to 12 months, expediting the cash flow for the bank during financial crisis.
• Analysed the shift in consumer spending patterns (essential & non-essential spends) during Euro Zone Crisis; flagged a red signal to the authorization team for luxury transactions.Achievements

Ms Artificial Intelligence & Machine Learning (Cs)

Summer School, Artificial Intelligence ( Computer Vision) , Machine Learning

Bachelor Of Engineering (Be), Electronics And Communication