Swayam

Data Analyst, Data Scientist

Gandhi Engineering CollegeOdishafull_time, internship, freelance

Open to roles

PythonSQLMachine Learningpandas

Student Expense Tracker

College students in India manage money across UPI, canteens, travel, and recharges — but rarely know where it all goes. Existing finance apps need internet, account creation, and often push ads. Most students on tight budgets don't always have reliable internet or a smartphone app they trust. There was a clear need for a simple, offline, zero-setup tool that runs straight from the terminal with no install. The challenge: build something genuinely useful that a first-year student could run and benefit from on day one. Process I started by reading the brief carefully and mapping out the data structure before writing any code — a list of dictionaries, where each expense holds a date, amount, category, and note. I built each feature as a standalone function so they could be tested independently. The budget warning was added last as a cross-cutting check across all menu options. I paid close attention to input validation since CLI tools break easily on bad input — wrong dates, non-numeric amounts, and invalid categories are all handled gracefully. Finally, I pre-loaded three realistic Indian-context sample expenses so reviewers see working output the moment they run it, without adding anything themselves. Results The program runs without errors on all tested inputs. A student can type python expense_tracker.py and instantly see a working menu with sample data. Adding an expense takes under 10 seconds. The category bar chart makes it immediately clear where money is going. The budget warning gives a proactive heads-up at 80% before overspending happens. Bad inputs — wrong dates, letters instead of numbers, unknown categories — are all caught and explained without crashing. All five menu options work and return cleanly to the menu loop. The experience feels complete, not like a rough student project. Reflection The program runs without errors on all tested inputs. A student can type python expense_tracker.py and instantly see a working menu with sample data. Adding an expense takes under 10 seconds. The category bar chart makes it immediately clear where money is going. The budget warning gives a proactive heads-up at 80% before overspending happens. Bad inputs — wrong dates, letters instead of numbers, unknown categories — are all caught and explained without crashing. All five menu options work and return cleanly to the menu loop. The experience feels complete, not like a rough student project.

2 media filesView

IPL Match Intelligence: Data-Driven Analysis of Winning Patterns in T20 Cricket

Cricket discussions are often driven by assumptions rather than data-backed insights. This project aimed to identify the real factors influencing IPL match outcomes using ball-by-ball analysis. Instead of focusing only on basic statistics, the analysis explored toss impact, match phases, batting consistency, bowling performance, and venue conditions. The project also focused on transforming a large IPL dataset into simple, understandable insights using visualization and analytical storytelling. Process The project started with exploring and understanding the IPL ball-by-ball dataset containing match details, player statistics, scoring patterns, toss results, and venue information across multiple seasons. After identifying the important columns, the dataset was cleaned by handling missing values, removing unnecessary data, and standardizing column names. Exploratory data analysis was then performed to study match phases, batting consistency, bowling performance, and scoring trends. Visualizations were created using Pandas, Matplotlib, and Seaborn to transform raw IPL data into clear analytical insights and storytelling-driven charts. Results The project revealed several important insights about IPL match outcomes using ball-by-ball data analysis. One of the biggest findings was that strong death-over performance influences victories more than toss advantage. The analysis also showed that teams maintaining momentum during middle and final overs perform more consistently across seasons. Venue conditions, batting stability, and disciplined bowling also played major roles in match results. Overall, the project transformed raw IPL data into meaningful sports analytics insights. Reflection If I were to improve this project further, I would focus more on predictive analytics and interactive dashboards instead of relying mainly on descriptive analysis. While the current project successfully explains IPL match patterns and scoring trends, adding machine learning models could help predict match outcomes and player performance more effectively. I would also improve visualization design by creating cleaner and more interactive charts. Additionally, deeper player-level analysis and live data integration could make the project more dynamic and strategically valuable.

7 media filesView

This is Swayam’s work on Wooble.

Build a profile that shows what you can do — and share it anywhere.

Build yours