Summary
Overview
Work History
Education
Skills
Accomplishments
Timeline
Generic

Govind Saria

Bengaluru

Summary

Accomplished Data Scientist with 10 years of expertise in leading data science initiatives and generating actionable insights for complex business challenges. Skilled in deep learning, Generative AI, and machine learning, with hands-on proficiency in statistical analysis and predictive modeling. Passionate about solving business problems through data. Recognized for strong communication, leadership, and collaboration in delivering innovative, business-aligned solutions.

Overview

10
10
years of professional experience

Work History

Staff Data Scientist

Walmart Global Tech
02.2020 - Current
  • Led a team of Data Scientists and MLEs in developing a patented multi-market solution for HTS code classification, easing cross-border trade for Marketplace sellers. Implemented an ensemble of NLP and computer vision models to classify items to their respective HS codes to calculate the duty amount required at customs. The solution has influenced GMV around 50M$ in last one year in limited release and expected to do 300M$ in next 1 year and around 2B$ in next 4 years upon global rollout.
  • Led a GenAI project on HTS classification, using zero-shot learning to generate training data for missing categories, which in turn enhanced model accuracy via a feedback loop with minimal labeled data. We also leveraged GenAI to extract product attributes and key context, enabling ops and compliance teams to make informed decisions, especially when deep learning models had low confidence scores. This streamlined classification and decision-making processes.
  • Managed, reviewed, and oversaw the Marketplace Data Science team’s design and development work on recommendation models for the Review Accelerator Program under multiple initiatives to help sellers boost reviews on their items. The solution was crafted in line with business requirements and industry best practices. The solutions have helped us to add around 1M additional reviews per month consistently over past one year.
  • Acted as data science SME and provided thought leadership to Seller Risk team to design the methodology to combat payment frauds, copyright/trademark violations and seller-seller collusion which brought Walmart Marketplace payment losses from 5bps to under 3bps and improved customer experience. Designed re-usable features like Brand Quality Score, Trust Score which can be used across Marketplace in various solutions to combat brand risk.
  • Demonstrated thought leadership by driving operational excellence within the team, streamlining best coding practices, standardizing GIT workflows, and implementing consistent CI/CD pipelines. Led efforts to track code coverage across projects, ensuring high-quality deliverables and fostering a culture of continuous improvement.
  • Organized and led knowledge-sharing sessions across Marketplace, bringing together multiple teams for joint discussions on diverse tech topics, fostering collaboration and driving collective growth within the organization.
  • One of the founding members of the Marketplace Data Science team, and helped it grow from 5 members to a 30-member team over 4+ years by conducting 200+ interviews. Managed interns for 3 consecutive years to support campus partnerships.

Data Scientist

Numerify (Acquired by Digital.ai)
08.2018 - 02.2020
  • Lead a team of data scientists to develop flagship data science solutions like Change Risk Prediction, Incident
    prediction, Change-Incident Linkage for the company.
  • Deployed Change Risk prediction for 10+ Fortune 500 companies netting 25M+ revenue from the solution.
  • Helped team grow by series of referrals and interviews

Team Lead

Nash Ventures (Closed)
01.2018 - 05.2018
  • Led a team of software engineers and analysts to develop infrastructure for algorithmic trading in python
  • Developed Statistical and Machine learning based algorithms for trading equities, metals and cryptocurrencies

Data Scientist

Paysense (Acquired by PayU)
05.2017 - 12.2017
  • Responsible for Credit default risk strategies and analysis.
  • Developed Machine Learning models to identify probable future credit defaulters and non-defaulters reducing the overall default rate by 5bps.
  • Developed CNN based model to classify front and back aadhaar images from non-aadhaar ones which
    combined with OCR was used to do first level document verification on app enabling 5-minute loan approval.

Member of Technical Staff

VMware
08.2014 - 04.2017
  • Worked as back-end developer on VMware cloud product Hybridity Cloud Manager (HCM) which facilitates live migration of Virtual Machines as well as network extension instantiations.
  • Collaborated with Other teams for Data Analysis and IOT discovery projects under CTO office

Education

Bachelor of Technology - Electrical Engineering

IIT Kanpur
Kanpur, India

Master of Science - Data Science

University of Arizona
Arizona, USA

Skills

  • Statistical Methodology: Hypothesis testing, ANOVA, regression analysis, time-series forecasting, and experimental design
  • Machine Learning: Expertise in both classical and modern machine learning algorithms, including supervised learning (eg, linear regression, decision trees, random forests, gradient boosting), unsupervised learning (eg, k-means, PCA, hierarchical clustering), and ensemble methods Familiarity with model evaluation metrics, feature engineering, hyper-parameter tuning, and cross-validation techniques
  • Deep Learning: Neural network architectures, such as CNNs for image-related tasks, Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks for sequence data, and transformer-based models (eg, BERT, GPT) for advanced NLP tasks Proficiency in using deep learning frameworks like TensorFlow, PyTorch, or Keras for developing and deploying models at scale
  • NLP: Expertise in processing and analyzing large-scale textual data, including techniques such as tokenization, stemming, lemmatization, and the use of word embeddings (eg, Word2Vec, GloVe) Advanced skills in building and fine-tuning language models for tasks like text classification, sentiment analysis, named entity recognition (NER), and topic modeling using tools like NLTK, SpaCy, and Hugging Face Transformers
  • GenAI: Variational Autoencoders (VAEs), Diffusion models Techniques like LORA, QLORA, RAG etc

Accomplishments

  • Filed a patent on Multi-Market HTS code Classification in Walmart
  • Lead author for multiple papers and posters in Walmart's Internal and External, Conference (AI Summit)
  • Won VMware All India Hackathon in December 2015

Timeline

Staff Data Scientist

Walmart Global Tech
02.2020 - Current

Data Scientist

Numerify (Acquired by Digital.ai)
08.2018 - 02.2020

Team Lead

Nash Ventures (Closed)
01.2018 - 05.2018

Data Scientist

Paysense (Acquired by PayU)
05.2017 - 12.2017

Member of Technical Staff

VMware
08.2014 - 04.2017

Bachelor of Technology - Electrical Engineering

IIT Kanpur

Master of Science - Data Science

University of Arizona

Similar Profiles

CREATE PROFILE
Govind Saria