harishalakshanwarnakulasuriya.tech

harishalakshanwarnakulasuriya.tech When exploring most websites, you’ll notice many common features. But here, we are different.

Our platform stands apart from other websites and products through innovation, precision, and a commitment to delivering a truly unique experience.

05/03/2026

🚀 Major Project Update: Harisha's Mission Control V3.0

🌐 Live Now:
https://mission.harishacrypto.xyz

A massive architectural upgrade has just landed.

Mission Control is no longer just a dashboard — it’s now a high-performance AI productivity ecosystem designed for speed, scalability, and stability.

This update restructures the core navigation and styling engine, allowing the platform to seamlessly support 25+ AI and productivity tools.

⚡ Key Upgrades

🔥 Lightning-Fast Navigation
Traditional routing (react-router-dom) has been replaced with a state-driven dynamic component loader.
Result: Instant tool switching with zero page reloads.

🎨 Bulletproof Theming System
Integrated Material-UI ThemeProvider with safe-fallback hex values to eliminate UI crashes and ensure consistent rendering across all components.

🌙 Optimum Theme CSS Architecture
Thousands of lines of hardcoded CSS were transformed into dynamic CSS variables (:root) — making the platform cleaner, scalable, and ready for 1-click Dark Mode.

🔗 Perfect Component Synchronization
Core UI elements like SideDrawer, Header, Notifications, and UserMenu now communicate through a centralized state engine, ensuring flawless UI behavior.

🛠 Build & Dependency Optimization
Cleaned the entire dependency system, resolved Tailwind/PostCSS conflicts, and introduced cross-platform CI=false build rules for stable production deployments.

💡 Strategic Advantages

📈 Infinite Scalability
Adding new apps is now effortless.
Drop in a new tool — and the system auto-generates navigation and UI elements.

🧱 Rock-Solid Maintainability
Global design changes (like brand color updates) now require just one line of code.

🖥 Next-Gen User Experience
The platform now behaves like a native desktop application with fluid transitions between complex AI tools.

⚙️ Mission Control Vision

Building an AI-driven productivity ecosystem where developers, creators, and innovators can access powerful tools from a single intelligent dashboard.

The journey continues. 🚀










03/03/2026

🚀 New Project: ImageEditor Global – High-Performance Native Android Image Engine

I am excited to announce the completion of ImageEditor Global, a full-stack Android application designed to push the boundaries of mobile image processing. This project isn't just a standard photo editor; it is a demonstration of Native C++ (JNI/NDK) integration and Advanced Computer Vision techniques.

Live now: www.harishalakshanwarnakulasuriya.com

🛠️ Key Technical Features:

* PhD-Level Native Engine (C++ / JNI): Optimized for performance, I bypassed the JVM to implement a 2-Pass Gaussian Blur using custom convolution kernels and pointer arithmetic for zero-copy pixel manipulation.
* Interactive Layer Architecture: A custom-built rendering canvas that manages a dynamic stack of layers (Images, Text) using Multi-touch Matrix Mathematics (Scale, Rotate, Translate).
* Modern Camera Stack: Integrated CameraX API for high-resolution hardware capture and lifecycle-aware preview management.
* Hardware-Accelerated Filters: Real-time adjustments using *ColorMatrix transformations, achieving 60FPS fluid UI performance.
* Global Market Security: Full R8/ProGuard obfuscation to protect native logic and MediaStore API integration for Scoped Storage compliance (API 34+).

🎓 Key Engineering Takeaways:

1. Memory Management: Learned to "lock" bitmap pixels in the native layer to prevent GC overhead during heavy mathematical operations.
2. UX & Performance: Balanced heavy computational tasks with a smooth UI by offloading NDK processing to background threads while maintaining a modern API 31+ Splash Screen experience.
3. Mathematical Compositing: Implemented a flattening engine that renders complex transformation matrices into a single high-resolution export.

This project represents my deep dive into Android Systems Engineering and High-Performance Computing. I’m looking forward to applying these native processing skills to even more complex Computer Vision and AI challenges!

03/03/2026

🚀 New Project: ImageEditor Global – High-Performance Native Android Image Engine

I am excited to announce the completion of ImageEditor Global, a full-stack Android application designed to push the boundaries of mobile image processing. This project isn't just a standard photo editor; it is a demonstration of Native C++ (JNI/NDK) integration and Advanced Computer Vision techniques.

Live now: www.harishalakshanwarnakulasuriya.com

🛠️ Key Technical Features:

* PhD-Level Native Engine (C++ / JNI): Optimized for performance, I bypassed the JVM to implement a 2-Pass Gaussian Blur using custom convolution kernels and pointer arithmetic for zero-copy pixel manipulation.
* Interactive Layer Architecture: A custom-built rendering canvas that manages a dynamic stack of layers (Images, Text) using Multi-touch Matrix Mathematics (Scale, Rotate, Translate).
* Modern Camera Stack: Integrated CameraX API for high-resolution hardware capture and lifecycle-aware preview management.
* Hardware-Accelerated Filters: Real-time adjustments using *ColorMatrix transformations, achieving 60FPS fluid UI performance.
* Global Market Security: Full R8/ProGuard obfuscation to protect native logic and MediaStore API integration for Scoped Storage compliance (API 34+).

🎓 Key Engineering Takeaways:

1. Memory Management: Learned to "lock" bitmap pixels in the native layer to prevent GC overhead during heavy mathematical operations.
2. UX & Performance: Balanced heavy computational tasks with a smooth UI by offloading NDK processing to background threads while maintaining a modern API 31+ Splash Screen experience.
3. Mathematical Compositing: Implemented a flattening engine that renders complex transformation matrices into a single high-resolution export.

This project represents my deep dive into Android Systems Engineering and High-Performance Computing. I’m looking forward to applying these native processing skills to even more complex Computer Vision and AI challenges!

02/03/2026

Building an OS-Level AGI: Upgrading the Sinhala Voice Agent with Docker-Style Orchestration. 🚀

Beyond LLMs: Implementing Bayesian Utility Engines and System Telemetry in a Native Android Agent. 🧠

The "Agent Kernel" just got a massive upgrade. 🚀

I’ve been working on pushing the boundaries of what a mobile AI assistant can do—moving away from simple "if-else" logic toward a **Docker-style Orchestration** model.

The **Sinhala Voice Agent** is no longer just a calculator; it is now an integrated OS-layer agent that manages hardware, high-level mathematics, and cognitive strategy through a unified Omni-Cognitive Hypervisor.

What’s new in this V2 Upgrade?

🔹 Docker-style Node Orchestration: The Kernel now acts as a Host OS. It manages specialized "containers" of logic. Whether it’s a doctoral-level matrix inversion or a simple flashlight toggle, the system orchestrates threads to ensure high-performance ex*****on without system lag.

🔹 System Telemetry (Node Health): Borrowing from DevOps principles, the Agent now monitors its own "Host" (the phone). It tracks **Real-time RAM availability** and **Battery Health** to decide if it has the resources to run complex AGI workloads.

🔹 Bayesian Decision Core:The assistant doesn't just respond; it *thinks*. Using a Bayesian Utility Engine, it calculates the "utility score" of its next move—deciding whether to execute immediately, explain its reasoning (Tutor Mode), or request confirmation based on user trust levels and mood scores.

🔹 Hardware Interoperability:Deep integration with Android actuators. From WiFi orchestration and volume management to dynamic app launching—all handled via a native Sinhala/Singlish NLP interface.

Key Takeaways for Engineers:

✅ Separation of Concerns: The UI (Face), the Kernel (Host), and the Hypervisor (Logic) are fully decoupled.
✅ Neuro-Symbolic AI: Combining symbolic math parsing with heuristic neural weight adjustments.
✅ Performance First:Leveraging `WorkStealingPool` and `AtomicBoolean` for thread-safe, low-latency OS interrupts.

The goal isn't just to build an app—it’s to build a responsive, context-aware environment that speaks the user's language natively.

02/03/2026

Standard static portfolios are dead. So, I decided to build a living, breathing AI Ecosystem. 🚀

As I continue my M.Sc. in Computer Science at the University of Sri Jayewardenepura, I wanted to push the boundaries of distributed systems, edge computing, and user experience. I recently architected and deployed a fully autonomous, context-aware web application that literally reads human psychology and alters its own code in real-time.

This isn't just a React site—it's a Distributed Artificial Intelligence Agent powered by a Python/Docker mothership.

Live Now : www.harishalakshanwarnakulasuriya.com

Here is a look under the hood at the architecture:

🧠 1. The Edge AI Kernel (Local Symbolic Engine)
Running entirely in the browser, the custom React AI Kernel tracks spatial awareness, "dwell-time" (intent), and rapid click-rates (cognitive load/frustration). No hardcoded rules—it dynamically builds a neurological map of what the user cares about right now.

⚡ 2. Autonomous UI Actuation & Cinematic UX

Depth of Field Rendering: If the AI detects you are heavily researching my Cybersecurity or Post-Quantum Cryptography projects, it physically applies a heavy grayscale and camera-lens blur to irrelevant projects, forcing focus onto what you actually want to see.
* Frictionless Adaptation: If the "Frustration Sensor" detects erratic clicking, the engine instantly strips away heavy 1.5s cinematic animations, drops the custom cursors, and snaps the UI into a lightning-fast native state to reduce cognitive load.
* Dynamic Reordering: Skills and services literally glide and re-sort themselves via Framer Motion layouts to bubble your highest interests (like Cloud Infrastructure or Native Android) to the top.

📡 3. The Autonomic Nervous System & Cloud Telemetry
The Edge Agent silently batches user behavioral states and flushes telemetry to a Dockerized Python/Flask backend using background queues, ensuring zero UI lag.

🤖 4. Unsupervised Machine Learning (K-Means Clustering)
The Global Aggregation Node doesn’t just store data; it thinks. Using Scikit-Learn, it mathematically clusters visitors into personas, discovering hidden correlations (e.g., "Users exploring Mobile Java development also heavily index on OCI & Terraform deployments").

🔮 5. Predictive Recommendations (The Hive-Mind)
The React frontend reaches back to the Python ML model to generate Netflix-style, hyper-personalized project recommendations based on these global behavioral patterns. It even adds a glowing "🔥 Trending Globally" badge to projects that the internet is currently swarming.

Key Takeaway: We are entering an era where applications shouldn't just respond to clicks; they should anticipate intent, reduce friction autonomously, and learn from global telemetry without sacrificing edge performance.

I built this from my base in Malabe, but the architecture is designed to scale globally. If you're interested in distributed systems, machine learning, or next-generation UX, let’s connect!

Check out the live architecture below

16/02/2026

🚀 Bypassing Firewalls with Code: Introducing StealthVPN

I am thrilled to announce the official launch of StealthVPN – a custom-engineered, anti-censorship tunneling solution that bridges Android VpnService with raw Node.js socket programming.

🌐 Live Demo & Docs:
https://stealthvpn.harishacrypto.xyz

Building a VPN isn't just about moving data; it's about protocol translation. During development, I faced a critical engineering challenge: Android’s system proxy sends `HTTP CONNECT` headers, but standard tunnels expect `SOCKS5`.

💡 The Breakthrough:
I engineered a "Smart" Node.js Server that performs real-time packet sniffing on the first byte of incoming traffic.

* If it sees `0x05` → It handles the standard SOCKS5 handshake.
* If it sees `CONNECT` → It automatically parses the HTTP header and pipes the connection.
* All wrapped inside a custom *XOR-Obfuscated SSH Tunnel to defeat Deep Packet Inspection (DPI).

This project proves that with the right low-level socket logic, we can build tools that are secure, private, and resilient.

Check out the architecture :
👉 https://stealthvpn.harishacrypto.xyz

16/02/2026

We just fixed the biggest lie in influencer marketing. 🚀

For the last 5 years, we’ve been ranking creators based on the wrong numbers. Follower counts are vanity. Engagement rates can be gamed.

Agencies are still burning 40% of their budget on creators who have "reach" but zero resonance.

Today, I’m proud to unveil VibeRank Enterprise—the first Creator Intelligence Platform that moves beyond "How famous are they?" to "How well do they fit?"

Unlike legacy tools that just scrape profile stats, VibeRank uses predictive AI to analyze context, tone, and stability.

✨ The VibeRank Difference:
🔹 Tonal Resonance Scoring:Our semantic engine doesn't just read captions; it "gets" the joke. It distinguishes between *wholesome*, *sarcastic*, and *corporate-safe* creators instantly.
🔹 Burnout Risk Prediction:We analyze posting variance and sentiment to flag creators before they ghost your campaign. Stability is the new engagement.
🔹 Predictive Virality: Stop chasing last week’s trends. We identify "Trend Arbitrage" opportunities 14 days before they peak.
🔹 Ghost Attribution: Finally, measure the *invisible* lift. Correlate creator post times with local search volume spikes.

We aren't just building another database. We are building the Context Layer of the Creator Economy.

Key Takeaways :

1. The Shift:The industry is moving from Quantitative (Numbers) to Qualitative (Vibe/Fit). VibeRank leads this shift.
2. The Tech: We aren't a wrapper. We use custom predictive models and semantic analysis, not just basic APIs.
3. The Goal:To reduce campaign "miss rates" by ensuring tonal alignment before the contract is signed.

Start hiring for vibe, not just vanity.

👇 Link in comments for early access.

10/02/2026

🚀 Redefining Mobile Banking Security: Introducing SentinelVault Ultimate – An AI-Powered, Enterprise-Grade Financial Ecosystem.

I am thrilled to announce the completion of my latest major project, **SentinelVault Ultimate**.

As a Master of Science in Computer Science student, I wanted to go beyond standard banking apps. My goal was to solve the critical gaps in modern fintech: Reactive Security (blocking fraud after it happens) and Static Banking (showing history instead of predicting the future).

SentinelVault is a full-stack, native Android banking solution built with a "Security First" architecture, integrating a Hybrid Payment Gateway tailored for global and local (Sri Lankan) needs.

💡 The "Missing" Features I Built:

🔹 Predictive Insolvency Engine (Python/Flask): Instead of just tracking expenses, the app uses Machine Learning to forecast financial health. It warns users before they run out of funds based on recurring bill patterns.
🔹 Hybrid Payment Gateway: A custom architecture merging PayPal (for International USD transfers) and PayHere (for Local LKR/Genie/EzCash), solving the cross-border payment friction in Sri Lanka.
🔹 "Duress Mode" Protocol: A unique security layer. If forced to unlock the app under threat, entering a specific "Panic PIN" opens a decoy wallet with a fake low balance, protecting the user's real assets.
🔹 Continuous Biometric Auth: Enterprise-grade security that re-authenticates users via fingerprint for every high-value transaction, not just at login.

Key Technical Implementations:
✅ Microservices Architecture: Decoupled the Core Banking logic (Node.js) from the Intelligence Layer (Python/Flask).
✅ Security: Implemented SSL Pinning to prevent Man-in-the-Middle (MitM) attacks and Root Detection to secure the client environment.
✅ AI-Driven Risk Scoring: A custom Random Forest model that scores every transaction in real-time. If the risk > 80%, the transaction is blocked *before* it hits the ledger.
✅ Global/Local Ledger: Engineered a multi-currency wallet system supporting simultaneous LKR (via PayHere) and USD (via PayPal) balances with real-time Forex conversion.

🛠 The Tech Stack:

* Mobile: Native Android (Java) with MVVM Architecture & SSL Pinning.
* Core Backend: Node.js (Express) with Rate Limiting & Helmet Security.
* Intelligence Layer: Python (Flask) with Scikit-Learn for Risk Scoring.
* Database: MongoDB (Encrypted Data-at-Rest).

This project represents a complete convergence of Mobile Development, Backend Engineering, and Applied AI. It’s not just about moving money; it’s about protecting it intelligently.

I’d love to hear your thoughts on the "Duress Mode" concept – is this the future of personal banking security? 👇

05/02/2026

RESEARCH PAPER

Automated Cloud Infrastructure Orchestration for Scalable Ride-Hailing Platforms: A Terraform-Based Approach on Oracle Cloud Infrastructure

Abstract
In the rapidly evolving domain of urban mobility, the reliability and scalability of backend infrastructure are critical determinants of a ride-hailing platform's success. This paper presents the architectural design and implementation of "Yaka Rider", a next-generation ride-hailing solution for Sri Lanka. We propose a cost-effective, high-availability cloud infrastructure utilizing Oracle Cloud Infrastructure (OCI) and Infrastructure as Code (IaC) principles. By leveraging Terraform for automated orchestration, we demonstrate a reproducible deployment strategy that integrates Oracle Linux 8, flexible Load Balancing, and Autonomous Databases. The proposed architecture addresses common challenges in startup environments, such as cost optimization, security via "Bring Your Own Key" (BYOK) protocols, and fault tolerance through redundancy.
Keywords— Cloud Computing, Terraform, Infrastructure as Code, Oracle Cloud Infrastructure, Ride-Hailing, DevOps, High Availability.

READ MORE:

https://www.academia.edu/164446866/Automated_Cloud_Infrastructure_Orchestration_for_Scalable_Ride_Hailing_Platforms_A_Terraform_Based_Approach_on_Oracle_Cloud_Infrastructure?source=swp_share

Address

Malabe
10115

Telephone

+94750788368

Website

, , , http

Alerts

Be the first to know and let us send you an email when harishalakshanwarnakulasuriya.tech posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The Business

Send a message to harishalakshanwarnakulasuriya.tech:

Shortcuts

Share

Category