ANISH SHAH

Enthusiastic & prolific result-oriented Data Scientist with a profound interest in playing with messy data and sleuthed big
chunks of data to produce insightful and valuable information, explained them intuitively, and provided impactful solutions
to derive business decisions.

Phone: (716) 587-1789 www.linkedin.com/in/anish23/

Email: anishh.shah23@gmail.com

Technical Competency
• Proficient in Python, R, Machine Learning, NLP, SQL, Spark, Statistics, and A/B Testing.
• Data, ML & AI: TensorFlow, Keras, Spark ML, DL4J, ND4J, Scikit-Learn, NumPy, SciPy, DEAP, Pandas,
LangChain, LangGraph, Spark NLP, Pyomo, PuLP, BERT, ELMo, Google Dialogflow CX, Playbooks, Stanford
CoreNLP, Rasa, NLTK, Spacy, OpenNLP, Polyglot, Gensim, Flair, Azure OpenAI.
• Big Data: Apache Hadoop, Apache Spark, Apache Kafka, Snowflake, Amazon Redshift.
• Deployment: Docker, Kubernetes, AWS Bedrock, AWS Lambda, AWS Sagemaker, AWS S3, Azure ML, Azure
Data Factory (ADF), Jenkins.
• Visualization Tools: Jupyter Notebook, PowerBI, Tableau.
• Database-RDBMS: MySQL, PostgreSQL, Oracle.
• Database-NoSQL: MongoDB, Dgraph, DynamoDB.
• OS & Cloud Platforms: Linux, Windows, Microsoft Azure, Amazon Web Services (AWS), Google Cloud Platform.
• Project Management: Git, Jira, SCRUM.

Experience

Exelon: Data Scientist October 2020 – Current

• Achieved an increase in efficiency of 70% by replacing the legacy mainframe system and developing multiple
production-level end-to-end ML pipelines on Microsoft Azure.
• Built a model to predict major network incidents with a model accuracy of 86%, resulting in approximately $1
million saved over one year.
• Built and deployed RAG pipelines using LangChain, Azure OpenAI, and FAISS to enable secure, context-
aware semantic search and document Q&A, improving retrieval precision by 35%.
• Spearheaded the design of a multi-agent orchestration framework using Python, LangChain, and LLMs to
simulate synthetic agents handling real-time user flows like fraud detection, account inquiries, and refund
disputes.
• Fine-tuned Large Language Models (LLMs) such as GPT and BERT on domain-specific datasets to build
customized NLP solutions for tasks like document classification, summarization, and entity recognition, improving
model accuracy and relevance by over 30%.
• Project Development: Designed and developed scalable production-level recommendation and prediction
systems leveraging Machine Learning, Deep Learning, Natural Language Processing, Statistical Modeling using
Python to solve real-world business problems.
• NLP (Natural Language Processing) Techniques: Built projects utilizing NLP knowledge including text mining,
regex, bag of words, TF-IDF, Word2Vec, encoder-decoder networks, attention, BERT, PCA, Bi-LSTMs, cosine
similarity, NER, and information extraction.
• Designed Responsible AI guardrails for NLP systems in a regulated utility environment, using keyword sensitivity
checks, bias detection modules, and human-in-the-loop escalation strategies.
• Agile Project Coordinator: Pitched machine learning ideas, showed exploratory data analysis (EDA) and
presented project demo to front desk business users; suggested, collected and synthesized business
requirements based on use cases, created an effective roadmap towards the deployment of a production-level
machine learning application.
• Data Analysis: Translating numbers into meaningful facts for businesses to help them make better business
decisions; Perform cleansing, manipulation, analysis, and visualization of client data.
• Data Reporting: Generated data visualization dashboards for reporting using Microsoft PowerBI for senior
management, which saved about 60 hours of manual work each month.
 • Model Evaluation: Measured model performance using Confusion Matrix, AUC- ROC curve; and identified
accuracy, precision, recall and F1 score using Confusion Matrix; used GridSearch to tune hyperparameters and
evaluate a model for each combination of algorithm parameters specified in a grid, increased accuracy by 5%

Altice USA: Data Scientist August 2019 – October 2020

• Developed and deployed a machine learning pipeline in Databricks to forecast product-level sales for retail
promotions using XGBoost, improving forecast accuracy by 15% and enabling optimized inventory allocation that
reduced overstock by 10%.
• Conducted data analysis and developed predictive models using Python and SQL, to drive business decisions
and optimize operations using AWS S3 and Amazon Redshift.
• Built machine learning models for customer segmentation, churn prediction, and product recommendations,
resulting in a 15% increase in customer retention AWS Sagemaker in Python.
• Collaborated with cross-functional teams, including marketing, finance, and engineering, to ensure alignment on
data-driven initiatives.
• Visualized data insights and communicate findings to both technical and non-technical stakeholders through
presentations and dashboards.
• Managed data collection, cleaning, and storage processes for various business units, resulting in a 20%
improvement in data accuracy.

Buffalo Sewer Authority: Data Scientist May 2018 – August 2018

• Built a multiple regression model with an 80% accuracy in Python and ArcGIS in one of the projects, which would
help the City of Buffalo in future Landscape and Urban Planning projects.

The Research Foundation of SUNY Buffalo, NY: Graduate Research Assistant May 2018 – December 2018
• Technical assistance using Data Science techniques under the guidance of RENEW (Research and Education in
Energy, Environment and Water) fellow Mr. Kevin J. Meindl for the planning, design, construction, maintenance,
and monitoring of the Buffalo Sewer Authority’s green infrastructure program.

Tech-Max: Data Analyst June 2016 – July 2017

• Executed analysis of large datasets using SQL and BI tools such as Tableau to provide strategic direction to the
enterprise business resulting in a 5% increase in sales.
• Developed statistical models to forecast inventory and procurement cycles and also, assisted in developing
internal tools for data analysis.
• Conducted cost and benefit analysis of new ideas, scrutinize and track customer behavior to identify trends and
unmet needs.

Education
Master of Science in Data Science February – 2019
University at Buffalo, The State University of New York
Courses: Data-Intensive Computing, Machine Learning, Data Mining, Probability, Computational Algebra,
Databases.

Bachelor of Engineering in Information Technology June – 2016

University of Pune, India
Courses: Distributed Systems, Cloud Computing, Data Structures, Design and Analysis of Algorithms.