Hi, I'm Keertika Agrawal
Aspiring Data Analyst | Transforming Data into Insights
Final-year Chemical Engineering student with a passion for data analysis, turning raw data into meaningful insights that drive business decisions.
About Me
I'm a final-year Chemical Engineering student with a strong passion for Data Analysis. My journey into data analytics stems from my fascination with uncovering patterns and insights that can transform business strategies.
I specialize in cleaning, analyzing, and visualizing complex datasets to extract actionable insights. With hands-on experience in real-world projects and a continuous learning mindset, I'm excited to contribute to data-driven decision-making in innovative organizations.
Goal-Oriented
Focused on delivering data-driven solutions that create business impact
Analytical Mindset
Strong problem-solving skills with attention to detail and patterns
Technical Expertise
Proficient in data analysis tools, visualization, and programming
Technical Skills
Data Analysis Tools
- ▹Excel / Google Sheets
- ▹Data Pre-Processing
Database
- ▹MySQL
- ▹Data Modeling
Programming
- ▹Python
- ▹Pandas, NumPy, SciPy
Data Science
- ▹Supervised & Unsupervised Machine Learning
- ▹Scikit-learn
Visualization
- ▹Microsoft Power BI
- ▹Matplotlib, Seaborn
Statistics
- ▹Statistical Tests & Hypothesis Testing
- ▹SciPy
Experience
Data Analyst Intern
Worked on real-world datasets in a fast-paced startup environment, contributing to data-driven business insights.
- Performed data collection, cleaning and preprocessing
- Created compelling data visualizations and interactive dashboards to communicate insights effectively
- Contributed actionable insights for strategic business decisions
Transformed raw industrial data into actionable insights for optimizing deodorization performance.
- Helped the production team quickly track key refining KPIs without checking multiple batch logs.
- Identified optimal temperature ranges and highlighted batches with low FFA removal or abnormal processing time.
- Revealed trends that helped improve process consistency, yield quality, and operational decision-making.
Featured Projects
Technologies Used
Key Highlights
- ▹Cleaned and preprocessed large datasets to enhance data quality
- ▹Cleaned and preprocessed a financial news dataset, handling missing values and visualizing patterns
- ▹Engineered features, creating market direction labels (Up/Down/Neutral) and extracting time based features
- ▹Built Classification Model (Random Forest, XGBoost) to predict market direction based on financial event attributes
- ▹Developed Regression Model (Random Forest, XGBoost) to estimate the magnitude of index change using event driven features
- ▹Compared model performance and selected optimal models based on metrics, accuracy score, confusion matrix, mean square error, and R2 Score
Impact
Provides predictive insights into market direction and volatility, helping illustrate how data-driven models can support financial analysis and investment decision-making.
Technologies Used
Key Highlights
- ▹Cleaned and preprocessed large datasets to enhance data quality
- ▹Applied Shapiro-Wilk test for normality assessment
- ▹Implemented Yeo-Johnson transformation for data normalization
- ▹Used Standard Scaling for feature standardization
- ▹Applied PCA for dimensionality reduction and improved model interpretability
- ▹Developed K-means clustering model to segment players into 4 performance categories
- ▹Helped identifying top-performing and supporting players
Impact
Successfully classified striker performance levels with improved model accuracy and helped identify top-performing players and supporting players through systematic preprocessing.
Technologies Used
Key Highlights
- ▹Visualized complex sales trends and KPIs through interactive dashboards
- ▹Identified 37% growth in total sales (+186.59K)
- ▹Discovered 36% increase in total orders
- ▹Revealed 14% rise in customer base
- ▹Analyzed category performance: Clothing (top-selling), Books (least-selling)
- ▹Identified Electronics as the fastest-growing segment
- ▹Discovered seasonal patterns: peak sales in June (70K), lowest in July (44K)
Impact
Provided actionable insights that helped identify growth opportunities, optimize inventory for high-performing categories, and understand seasonal sales patterns.
Get In Touch
I'm currently open to data analyst opportunities and collaborations. Feel free to reach out if you'd like to discuss data analytics, projects, or potential opportunities!