Product-Focused Full-Stack Software Engineer — building production-ready SaaS & AI systems.

"AutomateLanka is a professional, self-hosted automation hub that bridges the gap between intelligent AI discovery and high-fidelity workflow execution. It offers a robust SaaS infrastructure for managing complex business processes at scale, featuring an intuitive node-based visual drag-and-drop canvas (XYFlow) and intelligent intent analysis powered by local ML models."

"JarvisX V2 is a custom-trained AI assistant built by fine-tuning the Mistral-7B model using 137,300 domain-specific examples across 7 operational modes. The system integrates engineering automation, system monitoring, business workflows, creative tools, and bilingual conversational intelligence. It supports both cloud and local deployment using Hugging Face, GGUF, and Ollama for efficient and scalable inference."

"--- title: SmartStore SaaS slug: smartstore-saas tagline: AI-Powered E-commerce Automation Platform. status: Active startDate: 2026-02-19 featured: true --- # SmartStore SaaS ## Basic Info - **Project Title:** SmartStore SaaS - **Slug:** smartstore-saas - **Tagline:** AI-Powered E-commerce Automation Platform. - **Status:** Active - **Start Date:** 2026-02-19 - **Featured Project:** Yes ## Demo & Media - **Hero Image URL:** https://... - **Repo URL (Manual):** - https://github.com/AsithaLKonara/SmartStoreSaaS-Mono - **Deployment URL:** https://smart-store-saas-demo.vercel.app/ - **Vercel Project ID / Name:** smarts tor e-saas - **GitHub Repo (owner/repo):** AsithaLKonara/SmartStoreSaaS-Mono ## Full Content - **Summary / Overview:** Architected and built a production-ready, multi-channel SaaS e-commerce platform with AI-powered recommendations and business intelligence. Actively used local AI models (Ollama) alongside OpenAI for business automation. - **Technologies:** Next.js 14, TypeScript, PostgreSQL, Redis, Prisma, Docker, Stripe, NextAuth, Tailwind CSS, Playwright, Jest, Supabase - **Problem:** E-commerce businesses lacked affordable, integrated AI recommendations and business intelligence tools. They struggled with complex text-based ordering (like WhatsApp), multi-tenant data isolation, and scattered payment gateways. - **Contribution:** Architected and built a production-ready, multi-channel SaaS platform with AI-powered recommendations, natural language order parsing, and an integrated real-time analytics engine. - **Impact:** Delivered a scalable backend, integrated payments, and AI-driven insights for store owners, supporting automation across inventory, financials, and customer support. ## Challenges - **Data Isolation in Multi-Tenant SaaS:** Ensuring strict boundaries so that each tenant's data (products, orders, customers) remains secure and isolated without performance degradation. - **Text-to-Order AI Parsing:** Converting unstructured natural language (e.g., WhatsApp messages) into structured order formats efficiently and accurately. - **Multi-Gateway Payment Integration:** Handling simultaneous support for local and international gateways (Stripe, PayPal, PayHere, LankaQR) along with a dynamic currency exchange engine. - **Real-Time System Monitoring and Telemetry:** Integrating smooth real-time telemetry ingestion and alert monitoring across a Next.js and Redis stack. ## Solutions - **Strict Database Scoping:** Implemented rigid database scoping at the Prisma ORM level by linking all core records directly to an `organizationId` with robust Role-Based Access Control (RBAC). - **Hybrid AI Engine:** Deployed an NLP engine leveraging both local AI models (Ollama) and OpenAI to handle text-to-order requests and smart customer segmentation efficiently. - **Unified Payment Architecture:** Abstracted payment flows into a multi-gateway processor, providing a seamless checkout experience regardless of the user's selected provider or currency. - **WebSocket & Redis Integration:** Built a robust event-driven system using Redis and WebSockets to deliver real-time operational intelligence and telemetry updates directly to the client interfaces. ## Outcomes - **Enhanced Operational Efficiency:** Automated order creation from WhatsApp, drastically reducing manual data entry for store owners. - **Scalable Architecture Deployed:** Successfully launched a high-performance orchestration engine capable of scaling with multiple tenants and managing large product catalogs. - **Improved Analytics & Insights:** Empowered organizations with an AI Financial Controller and Inventory Autopilot, enabling better business decisions and automated reordering. ## Case Study SmartStore SaaS is an advanced, AI-driven multi-channel commerce platform engineered to overcome the common shortfalls of traditional e-commerce solutions. By combining the power of Next.js 14, PostgreSQL, Redis, and a sophisticated local/cloud AI architecture, it transforms how businesses interact with clients and manage internal operations. The application introduces a robust multi-tenant design ensuring complete data isolation through organizational scopes. Rather than typical form-based entries, it embraces modern conversational commerce with its Text-to-Order NLP parser, allowing orders out of completely unstructured text—directly bridging platforms like WhatsApp with the core database. Smart integration of IoT telemetry metrics and dynamic multi-gateway pricing architectures ensure that store administrators always have a real-time, comprehensive view of their operations. SmartStore SaaS successfully demonstrates how modern edge computing and AI can automate standard workflows into intelligent business processes."