Sankalp Biswal

About Me

I'm a data scientist with expertise in Machine Learning, Statistical Analysis, Predictive Modeling, and Data Visualization. With over 2.5 years of experience developing predictive algorithms, conducting statistical testing, building interactive dashboards, and creating intelligent systems using RAG and LLMs across cloud platforms, I excel at transforming complex datasets into actionable business insights. My background from consulting and data science positions me with the right balance of client-facing and technical skills, enabling me to bridge the gap between advanced analytics and strategic business value using Python, SQL, and machine learning frameworks.

Programming

Python SQL R JavaScript

Machine Learning

scikit-learn PyTorch TensorFlow Keras LangChain MLflow

Cloud & Tools

AWS Azure GCP Docker Snowflake Databricks

Analytics

Tableau Power BI Pandas NumPy A/B Testing Statistics

Experience & Education

Work Exp

Data Scientist Intern

Allagash Brewing Company, Boston, MA

Jan 2025 - Mar 2025

Developed real-time truck recommendation Streamlit app reducing shipping costs by 10% ($150K annually).

Work Exp

Data Scientist Intern - Conversational AI

DataWorksAI, Boston, MA

Apr 2024 - Jun 2024

Built and deployed RAG-based LLM chatbot for university students.

Education

Master's in Data Analytics (Artificial Machine Intelligence)

Northeastern University, Boston, MA

Sep 2023 - Mar 2025

GPA: 4.00/4.00. Specialized in AI System Technologies, LLM+RAG, Predictive Analytics, and Healthcare Data Applications.

Work Exp

Data Scientist

Falcon Cables, New Delhi, India

Feb 2022 - Aug 2023

Designed ETL workflows and Power BI dashboards improving efficiency by 20%.

Work Exp

Analyst - Technology Consulting

Ernst & Young LLP, Gurugram, India

Jun 2021 - Jan 2022

Researched NBFCs, shaped $11M transformation.

Education

Bachelor of Technology in Computer Science

Jaypee Institute of Information Technology, Delhi, India

Jul 2017 - Aug 2021

CGPA: 7.3/10. Co-authored research paper on diabetic retinopathy detection.

Featured Projects

Text-to-SQL Agentic Healthcare Chatbot

Built a RAG-based LLM assistant using LangChain, Pinecone, and Snowflake to generate insights and visualizations from structured healthcare data via SQL generation with Gemini. Performed ETL on 2M+ rows of HIPAA-compliant NY SPARCS data.

LangChain Pinecone Snowflake Gemini ReAct Agent

Customer Churn Prediction with MLOps

Built a comprehensive churn prediction pipeline on 500K+ customer records using PySpark and AWS services with full MLOps implementation. Implemented CI/CD pipeline with GitHub Actions and performance monitoring in Grafana and MLflow.

PySpark AWS Glue SageMaker MLflow Grafana

Ad Campaign A/B Testing & Optimization

Applied causal inference via A/B testing on 580K+ users to identify statistically significant conversion drivers and optimize ad campaigns. Measured 43% relative uplift in conversion for test group.

A/B Testing Chi-Squared Mann-Whitney U Causal Inference Python

Chicago Traffic Crash Analysis

Analyzed 1.85M+ traffic crash records using Databricks, Delta Lake, and Azure Blob Storage. Merged datasets to uncover crash trends by weather, lighting, and causes, and built an interactive Databricks dashboard for safety insights.

Databricks Azure Blob Storage Delta Lake PySpark SQL

Student Math Score Prediction – End-to-End ML Deployment

Built an end-to-end ML pipeline (R² 0.88) with automated model selection across multiple regressors. Deployed a real-time app on AWS using Flask, Docker, and GitHub Actions CI/CD.

Python Scikit-learn XGBoost CatBoost Flask Docker AWS EC2 GitHub Actions

Logistics Optimization Suite – Allagash Brewing Company

Built a Logistics Optimization Suite with a Smart Truck Recommender WebApp (Streamlit + weather & routing APIs) and Power BI dashboards to cut freight costs, improve route efficiency, and optimize pallet utilization.

Streamlit Google Maps API OpenWeather API Python (pandas, requests) Power BI Logistics Optimization

Impact of the 2008 Financial Crisis on Heart-Disease Mortality (DiD)

Applied Difference-in-Differences to evaluate the effect of the 2008 financial crisis on heart-disease mortality, combining CDC, BLS, World Bank, and Yahoo Finance data to measure economic and health impacts.

Python Pandas Statsmodels Difference-in-Differences CDC BLS World Bank Yahoo Finance

Airbnb Booking & Pricing Dashboard

Developed an interactive Tableau dashboard to analyze Airbnb bookings and pricing trends across New York City. Visualized total bookings, average prices, and room-type distribution with dynamic filters and geographical heatmaps to support data-driven pricing and strategy decisions for property managers.

Tableau Data Visualization Interactive Dashboard Airbnb Data

Services Offered

Data Science

End-to-end data science solutions from data collection and preprocessing to advanced analytics, predictive modeling, and actionable insights that drive strategic business decisions.

Data Engineering

Designing and automating scalable data pipelines and infrastructure using cloud-native services and orchestration tools for both batch and streaming workflows.

Machine Learning

Building, validating, and deploying ML models to automate predictions and uncover patterns in large-scale structured datasets with domain-aligned business value.

Data Visualization

Developing insightful dashboards and visual stories that make complex data accessible and actionable for decision-makers across business domains.

Chatbot Development

Creating intelligent conversational AI solutions including RAG-based chatbots, voice AI agents, and automated customer service systems using cutting-edge LLM technologies.

N8N Workflow Development

Designing and implementing automated workflows and business process automation using N8N and other no-code/low-code platforms for operational efficiency.