"I don't just write code; I build solutions. As the Founder of Kuralara WebFlux and a Computer Engineering student, I bridge the gap between complex backend logic and intuitive user interfaces."
Motivated and dynamic, seeking opportunities to leverage technical expertise in AI/ML and Full-Stack development for impactful projects.
AI/ML, Full-Stack, Cloud
English, Tamil, Hindi
Leading a digital studio focused on modern web solutions and brand identity. Managing end-to-end client delivery.
Coordinating student technical publications, managing editorial workflow and technical documentation activities.
Leading AI project initiatives, mentoring students and guiding implementation of machine learning based projects.
Managed social media. Coordinated media teams for e-media campus initiatives.
Analyzed large datasets to extract actionable insights. Built predictive models and visualized trends.
Intelligent contactless railway inspection system using transformer-based semantic segmentation to detect rusted surfaces on fishplates. Calculates rust severity, identifies clean-metal regions and converts pixel coordinates into angular offsets to guide an ESP32-powered pan-tilt laser turret for precise QR encoding. Enables automated infrastructure monitoring and predictive maintenance.
Autonomous cybersecurity framework performing real-time threat detection, automated response and self-healing protection across cloud platforms, enterprise networks and APIs. Analyzes logs, traffic and user behaviour using adaptive AI models to detect zero-day attacks, isolate compromised resources and dynamically enforce Zero-Trust policies.
Developed a time-series forecasting framework to predict future power grid inertia under renewable energy integration. Implemented LSTM, GRU and BiLSTM models with multi-step forecasting to analyze long-term frequency stability. Results demonstrate improved prediction stability using LSTM for proactive grid reliability assessment in low-inertia power systems.
Proposed a measurement-based inertia estimation method derived from the swing equation using ROCOF extracted via Savitzky–Golay filtering. The framework provides physically interpretable real-time inertia monitoring without training data and demonstrates robust estimation across multiple operating conditions.
Open to collaborations, research discussions and innovative project opportunities.
Have an idea, collaboration or opportunity? Send a message.