Detail-oriented Data Analyst with 4 years of experience in extracting actionable insights from complex datasets to drive strategic decision-making and business growth. Expertise in utilizing advanced analytics, machine learning techniques, and data visualization tools to enhance data accuracy, streamline processes, and optimize resource allocation. Proven ability to develop predictive models and interactive dashboards that significantly improve operational efficiency and support key initiatives. Strong background in programming, statistical analysis, and project management, with a commitment to delivering high-quality, data-driven solutions. Skilled in Python, SQL, Tableau, and Azure, with a Master’s degree in Applied Data Science from Indiana University – Purdue University Indianapolis (IUPUI). Adept at collaborating with cross-functional teams and stakeholders to achieve organizational objectives and foster data-driven cultures.
Machine Learning Engineer May 2024 - Present
-
Currently contributing independently to the open-source Tree Tracker app, focusing on developing duplicate image detection and tree growth/update tracking functionalities.
-
Working on enhancing the app’s accuracy through the implementation of advanced image analysis algorithms.
-
Aiming to improve tree monitoring by creating mechanisms for detailed analysis of tree progress.
Project Data Analyst July 2022 – October 2022
-
Improved data accuracy and reporting efficiency by 25% through collecting, processing, and analyzing data from multiple sources for programs like "Rural Health Improvement" and "Women Empowerment" using Python and SQL.
-
Reduced decision-making time by 30% by developing and maintaining Tableau dashboards for initiatives such as "Education for All."
-
Increased targeted resource allocation by 20% by conducting statistical analysis to identify trends in service data for the "Clean Water Access" project.
-
Created and implemented business metrics to measure the effectiveness of various social service programs, leading to a 15% increase in program efficiency.
-
Prepared comprehensive reports and presentations using Excel and PowerPoint, effectively communicating key findings and recommendations, leading to a 10% increase in stakeholder engagement and program support.
Business Intelligence Analyst July 2019 – July 2022
-
Improved demand forecast accuracy by ~15% for a client using a Transformers-based package within Azure Databricks.
-
Enhanced supply chain visibility and decision-making through Power BI dashboards and Azure data pipelines, utilizing Databricks, DataFactory, and Synapse for Nokia’s data management team.
-
Worked collaboratively with data engineering teams to build and maintain data pipelines, ensuring high data quality and accessibility for analysis.
-
Reduced client ordering time by 70% by developing a backend framework and deploying it on Azure, enabling user rights management and bulk ordering while working directly with Nokia’s Executive Directors.
-
Designed and implemented business metrics and measurement systems, reducing post-deployment downtime by 17% through quality assurance tests using Selenium scripts, which allowed early detection of bugs in Nokia’s enterprise ordering platform.
-
Authored and presented detailed reports to Nokia's executive team to support strategic decision-making.
Programming Languages: Python, R, Java, HTML, CSS, JavaScript, C, C++, OOPs, SQL, Git
Machine Learning: Python (pandas, matplotlib, TensorFlow, NumPy, Spark, SpaCy, NLTK, Scikit-learn, Keras)
Data Science & Miscellaneous Technologies: : Jupyter Notebook, Git, Data cleaning, Data wrangling, Data visualization, Statistical and visual interpretation, Data Analytics, Statistical Analysis, Clustering, Time series Forecasting, Supervised ML, Unsupervised ML (Clustering, PCA), Generative AI, Large Language Models (LLM), Natural Language Processing (NLP), PySpark, Excel, Azure Azure (Databricks, Synapse, DataFactory), AWS (Sagemaker, AI Platform), Stata, Tableau, PowerBI, VSCode, Spyder
MS in Applied Data Science (May 2024) | Indiana University – Purdue University Indianapolis (IUPUI) Indianapolis, USA
Bachelor of Technology, Mechanical Engineering (June 2019) | Rajiv Gandhi Institute of Technology, Pampady Kottayam, India
Multimodal Reasoning via Chains of Thought (CoT) for ScienceQA dataset with varying Few-Shot Prompts
Implementing research to evaluate CoT models on ScienceQA with varied prompts. Utilizing UnifiedQA CoT, Multimodal CoT, T5, and VQA models to provide insights for NLP, AI, and education stakeholders.
RAG LangChain Chatbot for a Hospital System
Designed and deployed a RAG LangChain chatbot for a simulated hospital system, integrating custom data sources with Neo4j AuraDB. Developed comprehensive responses by fetching both structured and unstructured data. Ensured seamless accessibility for end-users by deploying the chatbot using FastAPI and Streamlit.
Forecasting Analysis on Rental Vacancy Rate in the United States
Analyzing and predicting US quarterly Rental Vacancy Rate from January 1956 to July 2023. Utilizing ARIMA, SARIMAX, SEASONAL NAÏVE, and PROPHET models for accurate forecasts. Evaluation based on MAE, MSE, and RMSE metrics, covering data from 2020 to the latest available.
Google - American Sign Language (ASL) Fingerspelling Recognition
Led a Kaggle competition project under Google for American Sign Language (ASL) Fingerspelling Recognition. Developed a deep learning model to predict ASL words from images of sign language gestures, achieving a 65% accuracy. Extended the model to perform real-time recognition using computer vision techniques.
Predicting Depression from patient data
Utilized NHANES and NAMCS datasets to extract patient health records and demographic information. Developed a machine learning model utilizing PHQ survey questions to predict mental health status of patients.