Surbhi Hirawat
Summary
Data scientist with 14+ years in AI/ML, NLP, and GenAI across fixed income, equity, and multimodal markets. Expertise in PyTorch/TensorFlow, AWS UltraClusters, RLHF, recommender systems, and information retrieval, delivering scalable solutions that optimize portfolios, detect anomalies, and drive measurable impact.
Overview
16
16
years of professional experience
Work History
Data Scientist
Elite Emante
07.2025 - Current
- Designed and deployed a marketing analytics assistant powered by large language models (LLMs) and Retrieval-Augmented Generation (RAG), enabling brand managers to query multi-channel campaign data in natural language and reducing manual reporting time by 40%.
- Built a RAG-based pipeline that ingested historical ad performance, customer feedback, and competitor insights to generate tailored marketing recommendations, driving a 15% improvement in campaign ROI through more precise audience targeting.
Data Scientist
Uber
04.2025 - 06.2025
- Led the EAS PAS migration with Python/SQL, hardening attribution pipelines and scaling to enterprise use; cut run times 30% and improved reliability.
- Built causal inference and ROAS optimization models (scikit-learn, TensorFlow) that drove measurable ROAS lift in pilot markets and guided budget shifts.
- Designed predictive analytics for churn, promo responsiveness, and behavior; sharpened targeting and reduced wasted impressions.
- Applied advanced causal inference and uplift models alongside traditional A/B testing to optimize promo responsiveness and personalization strategies, boosting targeting precision and ROI.
Executive Data Scientist
SportsBiz
09.2023 - 03.2025
- Directed cross-functional analytics initiatives; mentored teams and secured $2M to scale data platform and ML roadmap.
- Applied RLHF, self-supervised learning, and explainability methods to fine-tune LLMs for brand valuation, sentiment, and engagement measurement.
- Built GenAI pipelines with LangChain + RAG to process >100M documents, accelerating insights by 42% and reducing training time by 50%.
- Trained large-scale language and computer vision models using PyTorch, Hugging Face, Lightning, and VectorDBs, deployed on AWS UltraClusters for multimodal sports sponsorship analytics.
- Built and maintained large-scale ETL pipelines in Python and SQL, improving data reliability and reducing processing time by 30%, enabling faster analytics delivery.
Data Scientist
Mindhive LLC
06.2023 - 09.2023
- Created modular SQL and dbt-style transformations for reproducible analytics workflows, reducing ad hoc query burden for DS teams.
- Hardened attribution pipelines with QA checks and data validation frameworks, ensuring audit-ready and trustworthy datasets for decision-making.
Data Scientist Intern
RM Technotree
06.2022 - 08.2022
- Designed and implemented an NLP-based keyword tagging system using R and RStudio, simplifying podcast search and reducing manual content tagging efforts.
- Built an end-to-end data pipeline (ETL + EDA) and partnered with cross-functional teams to identify high-value sponsor leads, boosting lead generation efficiency through targeted data analytics.
Graduate Research Assistant
New York Institute of Technology
03.2022 - 05.2022
- Delivered interactive dashboards (Tableau) powered by optimized data models, empowering university to self-serve key metrics.
- Promoted innovation through prototyping, prioritizing model experiments, and presenting research findings with industry-aligned impact in collaborative academic settings.
Senior Data Scientist
Ames Developer
03.2019 - 09.2021
- Developed econometric models for fixed-income allocation and macro-risk forecasting, integrating ARIMA, Bayesian, and deep learning methods to improve portfolio optimization accuracy by 12% and support real-time market decisions.
- Built information retrieval pipelines to extract market signals from large volumes of FI research, integrating ML models with Spark for scalable performance.
- Developed Bayesian and DL-driven pricing frameworks, reducing pricing errors by 25% and supporting real-time macroeconomic risk assessment.
- Applied NLP and GenAI pipelines to process unstructured financial reports, extracting signals for allocation and risk monitoring.
Data Scientist
Inspira
12.2012 - 03.2019
- Designed recommender system-style models in Python/Spark to detect fraud and anomalies in equity markets, reducing false positives and fraudulent activity by 35%.
- Applied ML pipelines to equity trading and pricing data, integrating transaction signals with historical and macroeconomic features.
- Optimized model performance through A/B testing and causal inference, improving fraud detection and equity market strategy alignment.
- Built predictive models for pricing and risk reduction, combining market trend analysis with historical equity data.
- Standardized reporting pipelines across Qlik and Tableau, ensuring consistency in KPIs and enabling faster cross-team decision-making.
Hedge Funds-Investment
HSBC
02.2010 - 12.2012
- Managed multi-asset NAV/GAV hedge fund valuations and automated workflows for real-time investment reporting.
- Improved portfolio performance by 50% via TLM migration, providing financial analysis insights for strategic investments and leading cross-functional coordination.
Education
Master - Data Science
New York Institute of Technology
PG - Artificial Intelligence and Machine Learning
University of Texas
Skills
- Python, SQL, PowerBI, Tableau, Azure, R, scikit-learn, TensorFlow, CVAT, natural language processing, LLM, RAG GenAI, NoSQL, SAS, Hadoop, Teradata, RStudio, Matplotlib, Seaborn, ARIMA, LSTM, Java, Generative AI, AWS, Agile methodologies, Google Cloud
Timeline
Data Scientist
Elite Emante
07.2025 - Current
Data Scientist
Uber
04.2025 - 06.2025
Executive Data Scientist
SportsBiz
09.2023 - 03.2025
Data Scientist
Mindhive LLC
06.2023 - 09.2023
Data Scientist Intern
RM Technotree
06.2022 - 08.2022
Graduate Research Assistant
New York Institute of Technology
03.2022 - 05.2022
Senior Data Scientist
Ames Developer
03.2019 - 09.2021
Data Scientist
Inspira
12.2012 - 03.2019
Hedge Funds-Investment
HSBC
02.2010 - 12.2012
PG - Artificial Intelligence and Machine Learning
University of Texas
Master - Data Science
New York Institute of Technology