Roopkishor Singh profile photo 1 on Tal

Roopkishor Singh

Staff Data scientist at Toast | Ex-Goldman Sachs

Bangalore Urban

8+ years of work experience

Toast
Oct 20241 yr 11 mosPresent

Hybrid

Lead a high-performing engineering team delivering enterprise-grade GenAI, machine learning, and risk intelligence solutions. Act as a primary innovation driver, consistently pitching and scaling internal concepts to high-visibility executive buy-in. Spearheaded "Pulse," a first-of-its-kind Agentic AI Risk Intelligence Platform featuring autonomous agents that ingest merchant data to detect, score, and explain risk (Featured in The Economic Times). Architected a GenAI chargeback explanation system that drastically reduced dispute-challenge processing time. Led the fine-tuning and deployment of a lightweight LLM for transaction enrichment, driving a significant lift in NPS and an exponential increase in user adoption. Designed an fraud model that sharply reduced false positives, and built a chargeback-win model that protected millions in revenue for restaurant partners. Advanced Toast Lending capabilities by building LSTM/N-BEATS revenue forecasting models to improve accuracy and scale loan originations, and engineered long-term Probability of Default (PD) risk models.
Toast
Sep 20222 yrs 1 moSep 2024

Bengaluru, Karnataka, India

Goldman Sachs
May 20202 yrs 5 mosSep 2022

Bangalore Urban, Karnataka, India

Architected and deployed end-to-end machine learning solutions for personal finance management and customer engagement, scaling codebases to drive business integration. Personal Finance Infrastructure: Developed an intelligent Transaction Categorization engine using a Random Forest classifier, incorporating a human-in-the-loop stepwise feedback loop to continuously improve model accuracy based on user input. Predictive Income Modeling: Engineered an income estimation pipeline utilizing ARIMA time-series analysis to model historical paycheck data and forecast user income streams. Data Normalization: Designed robust string-matching and regex pipelines to clean and extract structured merchant names from raw, unstructured transaction payloads. Cloud Deployment & Architecture: Packaged and deployed all predictive models as production-ready APIs on AWS, ensuring high modularity, scalability, and seamless integration across multiple business units. Customer Engagement Systems: Built an intelligent notification engine and a Customer Lifetime Value (CLV) model driven by user app-interaction data to optimize retention and personal user targeting.
Goldman Sachs
Jun 20181 yr 11 mosApr 2020

Bangalore

Led the modernization of operational risk frameworks through automation, machine learning, and advanced predictive analytics. Pioneered ML in Risk Taxonomy: Single-handedly conceptualized and developed the organization’s first machine learning model to automate the classification of operational risk events into the L4 Risk Taxonomy, leveraging advanced NLP techniques to eliminate manual mapping. Automated Risk Governance: Spearheaded the development and end-to-end automation of critical Operational Risk Management Key Risk Indicators (KRIs), transitioning the business from reactive reporting to real-time risk mitigation. Financial Forecasting: Engineered an operational risk capital cost projection model utilizing time-series analysis to provide leadership with long-term financial foresight and strategic capital allocation strategies. Apple Card Risk Infrastructure: Designed and launched the foundational operational risk monitoring metrics and executive dashboards for the Apple Card portfolio, ensuring robust risk oversight for a high-profile product.
Bobble Keyboard
May 20173 mosJul 2017

Greater Delhi Area

User Feedback analysis System: My project basically focused on text analysis of the feedback given by the users. Sentiment analysis of the feedbacks using various libraries like NLTK, Stanford coreNLP and Text CNN. Feedback categorization of the negative feedbacks so as to use those feedbacks as Key Progress Indicators. Extracting the most demanded stickers and languages from the request feedbacks. Visualizing the trends of feedbacks over the years for various devices and versions using seaborn.
20162 yrs 5 mosMay 2018

IIT Roorkee

DSG, IITR is a group of data science enthusiast who aims towards developing the culture of data science in IIT Roorkee. Played an instrumental role in the establishment of the group Currently Leading a team of 22 people Conducted workshops and open meetings on Machine Learning and collaborated with industries for data science projects.

Roopkishor’s AI Stack

claude code
claude code
claude
claude
cursor
cursor
gemini
gemini
langchain
langchain
langgraph
langgraph

Education

Indian Institute of Technology, Roorkee

Bachelor’s Degree, Mechanical Engineering

Indian Institute of Technology, Roorkee

20144 yrs2018

High School

Tata Chemical DAV public School

199714 yrs2011

Join Roopkishor and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Roopkishor Singh on Tal

Headline

Staff Data scientist at Toast | Ex-Goldman Sachs

About

Staff Data Scientist with ~8 years of experience specializing in Agentic AI and Fintech Risk. Proven track record of spearheading cross-functional initiatives that deliver multi-million dollar revenue growth and enterprise-scale operational efficiencies.

Location

Bangalore Urban

AI stack

claude code

claude

cursor

gemini

langchain

langgraph

Work experience

Staff Data Scientist at Toast, Oct 2024 - Present · 1 yr 11 mos, Hybrid, Lead a high-performing engineering team delivering enterprise-grade GenAI, machine learning, and risk intelligence solutions. Act as a primary innovation driver, consistently pitching and scaling internal concepts to high-visibility executive buy-in. Spearheaded "Pulse," a first-of-its-kind Agentic AI Risk Intelligence Platform featuring autonomous agents that ingest merchant data to detect, score, and explain risk (Featured in The Economic Times). Architected a GenAI chargeback explanation system that drastically reduced dispute-challenge processing time. Led the fine-tuning and deployment of a lightweight LLM for transaction enrichment, driving a significant lift in NPS and an exponential increase in user adoption. Designed an fraud model that sharply reduced false positives, and built a chargeback-win model that protected millions in revenue for restaurant partners. Advanced Toast Lending capabilities by building LSTM/N-BEATS revenue forecasting models to improve accuracy and scale loan originations, and engineered long-term Probability of Default (PD) risk models.

Senior Data Scientist at Toast, Sep 2022 - Sep 2024 · 2 yrs 1 mo, Bengaluru, Karnataka, India

Data Scientist at Goldman Sachs, May 2020 - Sep 2022 · 2 yrs 5 mos, Bangalore Urban, Karnataka, India, Architected and deployed end-to-end machine learning solutions for personal finance management and customer engagement, scaling codebases to drive business integration. Personal Finance Infrastructure: Developed an intelligent Transaction Categorization engine using a Random Forest classifier, incorporating a human-in-the-loop stepwise feedback loop to continuously improve model accuracy based on user input. Predictive Income Modeling: Engineered an income estimation pipeline utilizing ARIMA time-series analysis to model historical paycheck data and forecast user income streams. Data Normalization: Designed robust string-matching and regex pipelines to clean and extract structured merchant names from raw, unstructured transaction payloads. Cloud Deployment & Architecture: Packaged and deployed all predictive models as production-ready APIs on AWS, ensuring high modularity, scalability, and seamless integration across multiple business units. Customer Engagement Systems: Built an intelligent notification engine and a Customer Lifetime Value (CLV) model driven by user app-interaction data to optimize retention and personal user targeting.

Risk Analyst at Goldman Sachs, Jun 2018 - Apr 2020 · 1 yr 11 mos, Bangalore, Led the modernization of operational risk frameworks through automation, machine learning, and advanced predictive analytics. Pioneered ML in Risk Taxonomy: Single-handedly conceptualized and developed the organization’s first machine learning model to automate the classification of operational risk events into the L4 Risk Taxonomy, leveraging advanced NLP techniques to eliminate manual mapping. Automated Risk Governance: Spearheaded the development and end-to-end automation of critical Operational Risk Management Key Risk Indicators (KRIs), transitioning the business from reactive reporting to real-time risk mitigation. Financial Forecasting: Engineered an operational risk capital cost projection model utilizing time-series analysis to provide leadership with long-term financial foresight and strategic capital allocation strategies. Apple Card Risk Infrastructure: Designed and launched the foundational operational risk monitoring metrics and executive dashboards for the Apple Card portfolio, ensuring robust risk oversight for a high-profile product.

Data Science Intern at Bobble Keyboard, May 2017 - Jul 2017 · 3 mos, Greater Delhi Area, User Feedback analysis System: My project basically focused on text analysis of the feedback given by the users. Sentiment analysis of the feedbacks using various libraries like NLTK, Stanford coreNLP and Text CNN. Feedback categorization of the negative feedbacks so as to use those feedbacks as Key Progress Indicators. Extracting the most demanded stickers and languages from the request feedbacks. Visualizing the trends of feedbacks over the years for various devices and versions using seaborn.

Senior Member at Data Science Group, IITR, 2016 - May 2018 · 2 yrs 5 mos, IIT Roorkee, DSG, IITR is a group of data science enthusiast who aims towards developing the culture of data science in IIT Roorkee. Played an instrumental role in the establishment of the group Currently Leading a team of 22 people Conducted workshops and open meetings on Machine Learning and collaborated with industries for data science projects.

Education

Bachelor’s Degree, Mechanical Engineering, Indian Institute of Technology, Roorkee

High School, Tata Chemical DAV public School