Shivam Mishra

[email protected] | 8789192996 | Bengaluru

https://github.com/ShivamMishra-codes | https://www.linkedin.com/in/shivam-mishra-07/

Data Scientist – II
Offering 4+ years of experience across Data Science, Data Analytics, Predictive Modeling, and
Machine Learning in the professional IT Industry

PROFILE
▪ Experience in implementation of Machine learning algorithms covering Classification, Regression,
Clustering, Anomaly Detection, Time series Forecasting, and Natural Language Processing
▪ Have good exposure to tools such as Jupyter Notebook/Lab, Google Collab, GitHub, Spyder, IBM
Watson, Power BI, Informatica, Heroku, MS Excel, Postman, TFS, and ALM

TECHNICALS
Tools Jupyter Notebook/Lab, Google Collab, GitHub, Spyder, IBM Watson, Power
BI, Informatica, Heroku, MS Excel, Postman, JIRA, TFS, and ALM
Packages Pandas, NumPy, SciPy, Keras, Scikit-learn, NLTK, spaCy, Matplotlib,
Seaborn, TensorFlow, statsmodel
Skills Python, Data Science, MySQL, Machine Learning, Predictive Modelling, Data
Wrangling, Exploratory Data Analysis, Data Visualization
PROFESSIONAL EXPERIENCE

Freelance Data Scientist

Upwork, Toptal, Fiverr 05/2023 to Present
Technologies Used: Python (V 3+), SQL (MySql), Data Science, N L P , Machine Learning

Key Accountabilities:
▪ Collaborated with multiple clients, developing custom fraud detection and data analysis solutions
tailored to specific business needs.
▪ Successfully developed and deployed advanced machine learning models for real-time fraud
detection, achieving up to 98% accuracy and reducing false positives by 30%.
▪ Designed and implemented risk scoring algorithms, categorizing transactions into various risk
levels.
▪ Conducted anomaly detection using Isolation Forest, One-Class SVM, and Autoencoders,
identifying fraudulent activities with high precision.
▪ Applied Natural Language Processing (NLP) techniques to analyze user behavior, detect fraud
patterns, and improve customer verification processes.
▪ Automated data wrangling and feature engineering processes, enhancing the efficiency of model
development and deployment.
▪ Conducted A/B testing to validate model performance and optimized fraud detection rules.
▪ Provided actionable insights and recommendations through detailed reports, helping clients
enhance their fraud prevention strategies.
▪ Developed and maintained strong client relationships, ensuring clear communication and
understanding of project requirements.
▪ Leveraged Supervised Algorithms including Light & Extreme Gradient Boost, Catboost, Lasso,
Random Forest, Ridge, Unsupervised Algorithms including Multivariate Gaussian probability,
Local Outlier Factor LOF, Robust Covariance (Elliptic Envelope)
▪ Used stacking and voting techniques improving accuracy by 15% with 10% volume increase.
▪ Developed and fine-tuned Generative AI models and Large Language Models (LLMs) for various
applications including content generation, text summarization, and conversational AI.
▪ Utilized state-of-the-art LLMs such as GPT-4 and BERT for natural language understanding, enhancing
client projects with advanced text analysis and generation capabilities.
▪ Conducted transfer learning on pre-trained models to customize them for specific client needs.
Verizon, India
Project: Enforce 03/2022 to 04/2023
Project Role: Data Scientist - II
Technologies: Python (V 3+), SQL (MySql), Data Science, Machine Learning

Key Accountabilities:
Developed and deployed advanced machine learning models for Fraud Detection in production
environment helping save $500 Million worth fraud team’s target using supervised, unsupervised,
semi supervised, graph, stacking & voting Machine learning approaches.

Below are the areas having experience on:

▪ Consulting & Designing data analytics solution for addressing process leakages, drive spending
& cost optimization.
▪ Implemented and deployed advanced ML Models in production environment detecting fraud
with more than 97%+ accuracy for inbound and 95%+ accuracy for offline orders.
▪ Risk Categorization of each incoming orders into low, medium, high and very high risk
categories using ML Model’s probability scores further helping in store representative
achieving almost 99% accuracy while order investigation.
▪ Minimized genuine customer friction by significantly reducing False Positive alarms
enhancing customer experience and retention with a jump of 20% in customer satisfaction.
▪ Implemented Auto Decisioning (Auto Release and Auto Cancel) functionality which further
contributes to projected $12 Million USD worth hard savings per year for which received
Spotlight Award for innovation from Verizon India CEO

Accenture Services Pvt Ltd, India.

Project: MyWizard 11/2020 to 03/2022
Project Role: Data Scientist - Analyst
Technologies: Python (V 3+), SQL (MySql), Data Science, Machine Learning

▪ Developed an AutoML application for AI model creation, deployment & automating data science life
cycle.
▪ Improved model accuracy by 50% for non-tree-based models through feature scaling,
hyperparameter tuning and skewness treatments.
▪ Implementation of Regression algorithms such as Random Forest, XG Boost, Ridge, Lasso,
Generalized linear model (GLM); Classification algorithms such as Random Forest, Gradient Boost,
XG Boost, Support Vector, Logistic Regression; Timeseries models such as ARIMA and SARIMA
▪ Increased R 2 accuracy score of GLM model by 20% using appropriate GLM family.
▪ Implementation of Unsupervised learning algorithms such as K-Means and DBSCAN clustering
algorithms and increased the silhouette score by 25%.
▪ Applied Natural Language Processing (NLP) techniques and implemented Similarity Analysis,
Text Summarization, and Sentiment Analysis.

Achievements/Certifications
▪ Received Spotlight Award FY22 for Innovation from Verizon India CEO for implementing Auto
Decision in Fraud Domain
▪ Received appreciation points & recognition from Manager on exemplary work and timely
deliverables in FY22 Q2 and in FY21.
▪ Completed IBM Data Science Professional Certificate specialization course offered by Coursera.
▪ Cleared Accenture’s Full-Stack Automation Engineer Certification in AI and Data Analytics

EDUCATION

Bachelor of Engineering (Information Science and Engineering) 08/2016 to 08/2020

Dayananda Sagar College of Engineering – 8.26 CGPA
Visvesvaraya Technological University, Bengaluru