Story Timeline

• Designed an ESP32-based trolley bot with wireless control capability.
• Highlights:
• - ESP32 WiFi control
• - Motor control and navigation
• - Lightweight firmware structure
• - Hardware-software integration

Hands-on embedded project emphasizing control logic and reliable WiFi command handling.

• Abinash S — Co-Creator
• Srivatsan N — Co-Creator
• HariPrasad — Co-Creator

• GitHub Repository
• ESP32

• Built the ICMAT conference site for SASTRA, focused on clear submission flow and accessible event information.
• Highlights:
• - Paper submission guidance and calls
• - Schedule and speaker section layout
• - Clean, responsive UI
• - Structured content sections

Designed to minimize confusion around submission requirements and keep deadlines and tracks easy to find.

• GitHub Repository

• Created a fest website for Theta2k23, highlighting event lists, announcements, and quick navigation.
• Highlights:
• - Compact, focused content
• - Mobile-first layout
• - Simple navigation flow
• - Event-centric sections

Prioritizes speed and clarity for event attendees on mobile.

• GitHub Repository

• Built a VS Code theme by forking Rosé Pine and customizing colors for clarity and long coding sessions.
• Highlights:
• - Rosé Pine base with tuned contrasts
• - Adjusted token colors for readability
• - Extension packaging and publishing
• - Consistent UI appearance

Theme focuses on readability and personalized color balance without losing the Rosé Pine character.

• GitHub Repository
• VS Code Marketplace (Hell Pine)

• Delivered a freelance website for a digital marketing company, emphasizing clear service presentation and client trust.
• Highlights:
• - Business-focused information architecture
• - Responsive layout and performance
• - Clean components and styling system
• - Client-ready delivery

Focused on clarity, credibility, and a solid handoff for client ownership.

• GitHub Repository

• Curated Neovim configuration focused on developer velocity.
• Highlights:
• - Modular Lua configuration
• - LSP and formatter integration
• - Fast startup and workflow shortcuts
• - Opinionated plugin setup

Config-first productivity project with an emphasis on speed and maintainability.

• GitHub Repository
• Neovim

• Hybrid IDS model combining Hierarchical Gated CNN (HGCNN) for spatial feature extraction and LSTM for temporal sequence modeling.
• Key Contributions:
• - HGCNN + LSTM end-to-end pipeline
• - Multi-class CPS attack detection
• - Evaluation on IIoT benchmark data
• - Clear ablation and error analysis

Notebook-driven experiments detailing preprocessing, feature engineering, and model training with reproducible metrics.

• Abinash S — Author
• Priyanga S — Author, Project Guide
• Srivatsan N — Author

• Research Repo (Notebooks)
• Project Report (PDF)
• UNSW-NB15 Dataset (UNSW)

• Hybrid CNN + LSTM IDS designed to detect malicious IIoT traffic with improved accuracy on benchmark data.
• Highlights:
• - UNSW-NB15 benchmark dataset
• - CNN for attack event detection
• - LSTM for temporal sequence modeling
• - Focus on lowering misclassification and false alarms

Notebook-driven workflow with clear dataset preprocessing, model training, and evaluation steps.

• Abinash S — Author
• Priyanga S — Author, Project Guide
• Srivatsan N — Author

• GitHub Repository
• Project Report (PDF)
• UNSW-NB15 Dataset (UNSW)

• Research implementation focused on cyber-physical system security.
• Highlights:
• - HGCNN + LSTM architecture
• - Sequence modeling for CPS data
• - Multi-class detection focus
• - Reproducible notebook workflow

Detailed model notebook includes preprocessing, training curves, and confusion-matrix evaluation.

• Abinash S — Author
• Priyanga S — Author, Project Guide

• GitHub Repository

• Research prototype targeting robust IDS performance in IIoT.
• Highlights:
• - Robust detection objective
• - Focus on IIoT threat resilience
• - Notebook-based workflow
• - Reproducible experimentation

Research artifact documents robustness-focused training and evaluation within Jupyter notebooks.

• Abinash S — Author
• Priyanga S — Author, Project Guide
• Srivatsan N — Author

• GitHub Repository

• Research project targeting APT detection using a tree-based genetic programming algorithm trained on highly imbalanced data (99:1).
• Highlights:
• - Genetic programming classifier (tree-based)
• - Handles extreme class imbalance (99:1)
• - Python-based experimentation pipeline
• - Reproducible training and evaluation workflow

Focuses on imbalance-aware training and GP-driven model evolution for minority attack detection.

• Abinash S — Author
• Priyanga S — Author, Project Guide
• Srivatsan N — Author

• GitHub Repository