Authenticating...
████████████████████████████ 100%
[ OK ] Identity verified
[ OK ] Secure connection established
[ OK ] Developer environment loaded
Welcome back, Krishni.
/\ krishni@github
/ \ ------------------------------
/\ \ OS: Windows
/ \ Host: VIT Chennai
/ ,, \ Kernel: Computer Science
/ | | \ Shell: Terminal
/_-'' ''-_\ Editor: VS Code
Languages: Java, Python, C,
C++, JavaScript, SQL
Focus: AI / ML, Data Science
Software Development
Open Source: GSSoC 2026
Interests: Astronomy, F1,
Japanese, Photography
Status: Always learning...
name: Krishni M R
education: B.Tech Computer Science Engineering
university: VIT Chennai
focus:
- Artificial Intelligence
- Machine Learning
- Data Science
- Software Engineering
- Backend Development
currently_exploring:
- Machine Learning applications
- Full-stack development
- Computer systems and networking
- Open-source collaboration
interests:
- Automotive technology and F1 telemetry
- Astronomy and astrophotography
- Space exploration
- Japanese language and culture
philosophy: "Learn. Build. Break. Debug. Repeat."
Five projects. Different problem domains. One goal: building things that work.
Repository: energypredictor_fds
An end-to-end machine learning project focused on understanding household energy consumption, forecasting future demand, and identifying potentially inefficient usage patterns using a dataset containing approximately 100,000 observations.
Inside the pipeline:
- Exploratory Data Analysis: Investigates household energy usage, temporal behaviour, environmental variables and feature relationships.
- Feature Engineering: Prepares temporal and environmental variables for model training and prediction.
- Model Benchmarking: Compares Linear Regression, Decision Tree, Random Forest, K-Nearest Neighbors and XGBoost.
- Dimensionality Reduction: Applies Principal Component Analysis (PCA) to explore feature structure.
- Consumption Pattern Analysis: Uses K-Means clustering to group usage patterns and investigate potential energy-waste behaviour.
- Time-Series Forecasting: Uses ARIMA to generate a 30-day energy consumption forecast.
- Prediction Interface: Integrates trained models into a Flask application with prediction functionality and forecast visualizations.
Core focus: Predictive modelling, time-series analysis and pattern discovery in a practical energy-monitoring application.
Repository: web_project_krishni
A full-stack classroom discovery and reservation platform designed to make room availability easier to inspect and manage. The application brings classroom information, scheduling and reservation workflows together.
System capabilities:
- Classroom Discovery: Centralized interface for exploring classroom availability and room information.
- Reservation Management: Classroom reservation workflows with scheduling conflict checks.
- Role-Based Workflows: Dedicated student, professor and administrative functionality.
- Timetable Integration: Scheduling information and availability-related operations.
- Real-Time Communication: WebSockets for live updates and notification workflows.
- Backend Architecture: Modular Express API routes for authentication, classrooms, reservations, timetables and other application features.
- Database Layer: Persistent storage for application and reservation data.
Core focus: Multi-role application design, scheduling logic, backend services, persistent data and real-time communication.
Repository: bck_smartcare_analytics
A hospital administration dashboard focused on turning operational information into a more accessible view of hospital activity, resource utilization and service efficiency.
What it brings together:
- Authentication & Access: Firebase Authentication with role-based administrative access.
- Operational Overview: Dashboard metrics for patient counts, average waiting times, bed availability and hospital efficiency.
- Data Visualization: Chart.js visualizations for operational patterns and performance indicators.
- Live Data Synchronization: Firebase Firestore integration for synchronizing dashboard data.
- Capacity Monitoring: Tracks waiting times and bed availability against configured thresholds.
- Patient Load Estimation: Uses historical capacity information and application-side estimation logic to explore next-day patient demand.
Core focus: Web development, cloud-backed data management and visual analytics for operational decision support.
Repository: simple_chat_v2
A computer-networks project exploring message exchange, protocol behaviour and bidirectional communication through an interactive protocol simulator and a live WebSocket implementation.
Protocol exploration:
- Two-Way Communication: Demonstrates communication initiated from both client and server sides.
- Message Identification: Uses message identifiers to associate messages with their corresponding acknowledgements.
- Acknowledgement Handling: Simulates ACK messages and tracks whether messages have been acknowledged.
- Timeout & Retransmission: Implements timeout handling and retry behaviour when acknowledgements are not received.
- Packet Loss Simulation: Simulates packet loss to observe how unreliable transmission affects communication.
- Event Logging: Displays message events, acknowledgements, timeouts and transmission failures.
Live implementation:
- Express server with WebSocket connections.
- Generates session codes for client connection.
- Supports a two-client session model.
- Relays messages between connected clients in both directions.
- Handles client disconnections and session resets.
Core focus: Communication protocols, acknowledgements, reliability, connection state and failure handling.
Repository: HousePricePrediction_MRKrishni
A supervised machine learning project exploring how housing attributes influence property prices and evaluating different regression algorithms on a housing dataset.
Modelling workflow:
- Data Preparation: Examines missing values, removes duplicate records and prepares features for modelling.
- Categorical Encoding: Converts categorical variables into machine-readable representations, including one-hot encoding.
- Model Development: Trains and compares Linear Regression, Random Forest Regressor and XGBoost Regressor.
- Performance Evaluation: Uses Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and the coefficient of determination (R²).
- Comparative Analysis: Evaluates model performance to understand the trade-offs between different regression approaches.
Observed model results in the project analysis:
| Model | R² Score |
|---|---|
| Linear Regression | 0.653 |
| XGBoost | 0.642 |
| Random Forest | 0.612 |
Core focus: Applying the complete regression workflow — from data cleaning and feature preparation to model comparison and evaluation.
I also contribute to collaborative open-source projects, working on features, algorithm visualizations, bug fixes and improvements.
| Repository | Contributions |
|---|---|
| AlgoScope | Stack behaviour, Tower of Hanoi, Prim's and Kruskal's algorithm visualizations |
| AlgoMate | Heap Sort, Floyd-Warshall, Counting Sort and stack capacity improvements |
| peer-learning | Portfolio functionality, Google OAuth and onboarding improvements |
| recode-website | Accessibility improvements and blog author-section styling |
| devtrack | Heatmap colour-scale improvements |
| sahidawa-india | Snooze tracker functionality and application improvements |
| 100_days_100_web_project | Sudoku and Dining Philosophers project fixes |
Collaborating, contributing and learning from the open-source community.
┌─────────────────────────────────────────────────────┐
│ BEYOND THE TERMINAL │
├─────────────────────────────────────────────────────┤
│ │
│ [01] SPACE │
│ Astronomy, astrophotography and exploration │
│ │
│ [02] AUTOMOTIVE │
│ Formula 1, telemetry and vehicle technology │
│ Red Bull Racing | Max Verstappen │
│ │
│ [03] JAPANESE │
│ Language learning and Japanese culture │
│ │
│ [04] PHOTOGRAPHY │
│ Capturing stories through a camera │
│ Instagram: framed.scene │
│ │
│ [05] ANIME & MUSIC │
│ One Piece, Attack on Titan, Haikyuu and more │
│ │
└─────────────────────────────────────────────────────┘
| Mission | Status |
|---|---|
| AI & Machine Learning | IN PROGRESS |
| Data Structures & Algorithms | IN PROGRESS |
| Open-Source Contributions | ACTIVE |
| Software Engineering | IN PROGRESS |
| Japanese Language | LEARNING |
| Astronomy & Astrophotography | EXPLORING |
┌──(krishni㉿github)-[~] └─$ echo "Thanks for stopping by!" Thanks for stopping by! ┌──(krishni㉿github)-[~] └─$ _
"The best code is written when the world is asleep." 🌙