I'm a software developer focused on building AI-powered applications, full-stack products, and intelligent software systems.

My work combines web engineering, backend development, database design, and AI integration. I'm also exploring machine learning, computer vision, and digital twins, with a long-term interest in connecting software intelligence to real-world systems.

I enjoy taking ideas from concept to working software: designing the architecture, building the application, integrating AI capabilities, and making the result useful in real business scenarios.

• Primary focus: AI Full-Stack Engineering
• AI interests: LLM-powered applications, machine learning, computer vision, and predictive maintenance
• Engineering interests: Digital twins, intelligent automation, 3D visualization, and software-hardware integration
• Product mindset: Building practical tools that solve real problems
• Location: Iran
• Portfolio: arshiapourfar.ir
• LinkedIn: Arshia Pourfar

AI SaaS · LLM Integration · Full-Stack Engineering

An AI chatbot platform designed to help businesses integrate intelligent customer support and product assistance into their websites.

Key features and development areas

• Embeddable website chat widget.
• Configurable LLM integration using customer-provided API keys.
• Product knowledge and business information workflows.
• Backend APIs and database integration.
• Workspace-oriented architecture and dashboard management.

Tech stack: Next.js · TypeScript · Prisma · PostgreSQL · LLM APIs

Status: Active development.

Live demo: Assist Chatbot

Source code: Private repository.

AI Email Automation · LLM Applications · Business Workflows

An AI-assisted email workflow designed to help process incoming leads, identify their intent, and prepare relevant response drafts.

Key development areas

• Lead intent classification.
• AI-generated email response drafts.
• Structured lead processing.
• Email and business workflow integration.

Tech stack: Next.js · TypeScript · Python · LLM APIs

Demo: AI Lead Responder

Digital Twin Simulation · Predictive Maintenance · Machine Learning

An engineering project exploring how physics-based simulations, operational data, and machine learning can support system monitoring and engineering decision-making.

Current development areas

• Wind-turbine behavior simulation using physics-based assumptions.
• Predictive-maintenance experiments using the AI4I 2020 dataset.
• Interactive dashboards for exploring system behavior and model outputs.
• Separate simulation and machine-learning workflows.

Tech stack: Python · Streamlit · Pandas · NumPy · Scikit-learn

Repository: Digital Twin Lab

Status: Under development.

Predictive-maintenance results are dataset-based experiments, not evidence of validation on industrial equipment. Simulation outputs depend on the assumptions and parameters used.

AI SaaS · 3D Visualization · Fashion Technology · Full-Stack Engineering

Fashion AI is a SaaS project focused on improving online clothing shopping through 3D visualization, body measurements, and AI-assisted clothing recommendations.

The long-term goal is to connect personalized sizing, garment presentation, and virtual try-on capabilities within an extensible platform for fashion retailers.

Core development areas

• 3D body visualization and parametric sizing concepts.
• Clothing size recommendations based on body measurements and garment data.
• Virtual try-on as an evolving product capability.
• API-oriented architecture for integration with clothing stores.
• Dashboard and store-management workflows.
• Exploration of computer vision and AI-assisted personalization.

Tech stack: Next.js · TypeScript · React · Three.js · React Three Fiber · PostgreSQL · AI/ML integrations

Status: Under development.

Source code: Private repository.

Advanced virtual try-on and computer vision capabilities should be treated as production features only when implemented and validated.

Machine Learning · E-commerce Analytics · Behavioral Data

An exploration of behavioral analytics and predictive modeling for e-commerce applications.

Areas of development

• User interaction event tracking and data summarization.
• Preparing behavioral and sales data for analysis.
• Exploring sales forecasting approaches.
• Investigating product recommendation methods.

Technologies: Python · FastAPI · TypeScript · PostgreSQL · Machine Learning

Forecasting accuracy and recommendation quality should be evaluated against documented results before being described as production-ready.

Full-Stack Development · End-to-End Engineering · Business Applications

An end-to-end e-commerce website independently designed and developed for a real business, covering the full development process from application architecture and database design to backend implementation and customer-facing functionality.

Key Contributions

• Independently developed the project from initial implementation to deployment.
• Built a responsive customer-facing storefront.
• Implemented backend APIs and database integration.
• Designed database models and application data workflows.
• Managed full-stack architecture and production deployment.

Tech Stack: Next.js · TypeScript · Prisma · PostgreSQL

Live Website: yekechin.com

Full-Stack Development · Business Digitalization · Restaurant Management

A full-stack digital platform designed to help cafés establish an online presence and manage essential business operations through a customer-facing website and an administrative dashboard.

Key Features & Project Scope

• Customer-Facing Website: Responsive café website for presenting business information and menu items.
• Digital Menu Management: Organize products, categories, pricing, and daily specials.
• Order Management: Backend support for handling and managing customer orders.
• Admin Dashboard: Centralized interface for managing café content and operational data.
• Analytics: Foundation for tracking business activity and generating operational insights.
• Authentication & Access Control: User authentication and role-based access management.

Tech Stack: Next.js · NestJS · TypeScript · Prisma · PostgreSQL · Neon

Status: Evolving project — ongoing development and testing.

Links: Live Demo

Feature descriptions should be kept aligned with the functionality implemented and verified in the current version.

Full-Stack Web Development · E-commerce

A web application project focused on building an online shopping experience with storefront functionality and administrative workflows.

Tech stack: Next.js · TypeScript · Node.js · PostgreSQL

Demo: Online Shop Platform

Frontend & Backend Development · Audio Applications

A music application featuring audio playback, favorites, persistent application data, and a responsive user interface.

Tech stack: React · Node.js · PostgreSQL

Demo: Music Streaming App

The technologies above reflect my development areas; the depth of hands-on experience varies by tool and project.

• Build end-to-end: Connect the frontend, backend, data layer, and AI components into usable applications.
• Solve real problems: Focus on user needs and business outcomes instead of adding technology for its own sake.
• Evaluate AI honestly: Distinguish prototypes from measured, reliable, production-ready systems.
• Design for iteration: Prefer clear architecture, modular components, and incremental improvements.
• Connect software with the physical world: Explore how simulation, sensor data, machine learning, and automation can work together.

• Improve and validate AI-powered business applications.
• Advance the Digital Twin Lab with clearer simulations, model evaluation, and engineering insights.
• Develop Fashion AI incrementally, focusing on 3D visualization and personalized sizing.
• Expand practical experience in machine learning and computer vision.
• Build reproducible project documentation and technical demonstrations.
• Explore the connection between AI software, sensor systems, and intelligent automation.

I'm interested in collaborating on AI-powered products, full-stack systems, machine learning applications, computer vision, and intelligent engineering projects.

Building useful software today. Exploring intelligent systems for tomorrow.