GENERATIVE AI
UI built on Streamlit, backend running on Ollama
Models available in Ollama library (including Deepseek, Llama3) can be pulled into local device to be used completely offline
Ability to create customized models by modifying the modelfile and creating custom bots based on open source models.
Built on langchain with Ollama backend to provide AI models
Interact with SQL Databases (currently uses SQLite) in natural language to query the database
Models available in Ollama library (including Deepseek, Llama3, Gemma) can be pulled into local device to be used completely offline
DATA ANALYTICS
Data sourced from Kaggle
Performed data cleaning and data manipulation on Python
Minor data transformation performed using Power Query
Dashboard visualization created on Power BI
Vehicle Sales Advertisement Analysis Dashboard
Data sourced from Kaggle
ETL performed on Power Query
Dashboard visualization of Vehicle Sales Advertisements in 2022 created in Power BI
Data obtained from Kaggle .
Performed exploratory data analysis, encoding categorical data and feature engineered the data.
Created a machine learning model to predict the price of apartments in Bangalore based on the modified data.
Cleaned the data, performed exploratory analysis and used GridSearchCV to conclude accuracy is higher for linear regression.
Bill Prediction in PySpark using Linear Regression
Explored the hotel bills data using PySpark on a Databricks cluster
Predicted the bill amount for the customer using Linear Regression on a vectorized column (using VectorAssembler)
Calculated the R2 value, mean squared and mean absolute errors.
HR Analytics Dashboard using Tableau
Created an interactive dashboard on Tableau (Tableau Public 2023) to analyse the employee attrition in a fictional organization
Using preprocessed data obtained from an internet based source
Sample Store Report using Power BI
Analysed the data and performed data manipulation on Power Query on sample data obtained on Kaggle.
Created a sample report on Power BI depicting profit for a sales in a store
Oracle Database Health Monitor
Analysed the different metrics of Oracle database relevant to the performance and availability of the database
Dataset generated (separate datasets for performance and availability) as comma-separated-values using ChatGPT
Oracle database Performance Dashboard and Availability Dashboards created on Power BI
ELECTRICAL ENGINEERING (CORE)
Adaptive Distance Relay For Renewables (B.Tech Major Project)
Implemented a MATLAB code and SIMULINK model for an adaptive distance relay using CFP41 numerical relay based on mathematical model proposed by:
K. Pradhan and G. Joos, "Adaptive Distance Relay Setting for Lines Connecting Wind Farms," in IEEE Transactions on Energy Conversion, vol. 22, no. 1, pp. 206-213, March 2007, doi: 10.1109/TEC.2006.889621.
Implemented a SIMULINK model for temperature control using a fan attached to a DC motor with power supply from a photovoltaic array along with a DC supply for backup power controlled by a IGBT switch connected to an ideal temperature sensor.
Implemented a SIMULINK model based on :
A. R. Prasad, P. D. Ziogas, and S. Manias, “An active power factor correction technique for three-phase diode rectifiers,” 20th Annual IEEE Power Electronics Specialists Conference. IEEE. doi: 10.1109/pesc.1989.48473)
OTHERS
Used the free API provided by GamerPower to extract info on giveaways and free offers for games.
Created a To Do App using Python Flask by referring relevant YouTube videos.
Created a falling balls game with the help of ChatGPT and relevant YouTube videos on pygame.
Created a word guessing game using pygame with the help of ChatGPT.