Full Stack Developer & AI Specialist - Crafting Intelligent Digital Solutions

Client: A mid-sized textile manufacturing company aiming to modernize their operations. Goal: Build an intelligent Product Lifecycle Management (PLM) system for the textile industry with real-time production tracking, fabric defect detection, and a direct marketplace. Challenge: Traditional systems lacked transparency, were expensive, and didn’t provide real-time quality control. There was also a need to automate manual processes like fabric inspection and order tracking. Solution: I led the development of “FiberFlow,” a scalable, AI-powered PLM platform. Using YOLOv5 and OpenCV, we implemented real-time fabric defect detection via both image uploads and live camera feeds. The platform also integrated camera-based garment progress tracking and a digital marketplace for manufacturers and buyers. Built with Django REST Framework (backend) and Next.js with TypeScript (frontend), it offered reliability, speed, and responsive design. Impact: The system helped improve production transparency, reduce defects and delays, and enhance buyer-seller trust. Rigorous unit testing across APIs, UI components, and AI models ensured platform stability. FiberFlow is now a benchmark for innovation in textile automation. Duration: 1-3 months · Budget: $1500 · Started 2024

Client: A music tech startup building an AI-first platform for artists, DJs, and producers. Goal: Develop a browser-based solution offering real-time audio tools like vocal remover, mastering, and chord detection using AI. Challenge: Deliver fast, accurate audio processing via deep learning models in a scalable web environment. The client needed seamless UX, secure data handling, and pricing integration. Solution: I led full-stack development using React for the frontend and Node.js with Python microservices for backend AI tasks. Integrated deep learning models (Spleeter-based via PyTorch) for real-time stem separation and FFmpeg for audio stream handling. Built cloud-native pipelines using AWS (Lambda, S3) to ensure scalability and speed. Stripe was added for usage-based billing, alongside secure user authentication and file handling. Impact: Delivered a sleek, responsive platform that empowers music creators to manipulate tracks directly from the browser. The system now supports thousands of sessions, high conversion rates, and seamless performance across devices. Duration: seven to thirty days · Budget: $1200 · Started 2024

Client: A university student team working on their Final Year Project (FYP). Goal: Develop an AI-driven academic hub to automate study planning, support assignment help, and track academic progress using LLMs and real-time analytics. Challenge: The client needed a scalable platform with personalized learning support, LLM-based interaction, and multi-user access—all while maintaining security and performance. Solution: I led the full-stack development using React (Next.js), Node.js, and PostgreSQL with Prisma ORM. I integrated Cohere’s LLM to provide automated assignment assistance and concept explanations through natural language prompts. The platform features interactive dashboards that use AI/ML insights to monitor study behavior, track tasks, and improve engagement. Clerk was used for secure authentication and role-based access. Deployment: Hosted on Microsoft Azure using a virtual machine to ensure speed and scalability. Outcome: The solution boosted student engagement and productivity, with over 85% of test users reporting improved focus and academic confidence. Duration: seven to thirty days · Budget: $1000 · Started 2024

Client: Barron Designs, a U.S.-based supplier of premium architectural products, needed a unified and scalable e-commerce platform to modernize their online sales process. Goal: Migrate four independent storefronts into a single, user-friendly BigCommerce-powered site with advanced product customization, better UX, and trade-specific features. Challenge: The existing ecosystem lacked flexibility and performance. The client required real-time product configuration (size, color, finish), dynamic pricing, SKU generation, and seamless backend automation — all while improving speed and mobile responsiveness. Solution: I led frontend and backend development, building a custom product configurator with real-time updates and integrated BigCommerce APIs to support dynamic SKU creation. I also developed a private backend app to automate product creation during checkout. The site included “Shop the Look” galleries, enhanced search, filtering, and B2B tools like sample kit requests and trade discounts. Impact: The redesigned platform improved engagement, reduced maintenance overhead, and enabled Barron to offer a personalized shopping experience to both retail and trade customers. Duration: seven to thirty days · Budget: $600 · Started 2025

I developed DevOpsLaunchpad, a complete learning platform built to teach modern DevOps practices through hands-on, real-world workflows. The goal was to create a professional, responsive web application offering curated educational resources and secure deployment. • Clean, responsive Next.js frontend with dark/light themes • Structured sections for DevOps tools, resources, articles • Downloadable cheat sheets for Git, Docker, Kubernetes • Secure, production-grade CI/CD pipeline using GitHub Actions • Automated testing, linting, code coverage with Codecov • Dependency vulnerability audits and secret scanning • Static analysis and container security integration • Automated Docker builds pushed to Docker Hub • Slack notifications for build and deployment events Challenges included integrating advanced security checks without slowing builds, managing secrets securely, and delivering an intuitive user experience. I solved these with modular pipeline design, GitHub Secrets, and Vercel environment management. The result is a polished, scalable platform for learners and professionals to master DevOps confidently. Duration: seven to thirty days · Budget: $550 · Started 2024

An advanced Autonomous AI Twitter Bot Dashboard designed as a complete scraper, responder, and analytics solution for social media management. • Goal: Automate Twitter engagement for a crypto community by scraping mentions, analyzing content, and posting intelligent, human-like replies. • Scraper Module: Continuously scans Twitter API for relevant $BTB mentions using advanced parsing logic. • Conversational AI: Uses OpenRouter-powered Large Language Models (LLMs) to generate context-aware, human-like responses. • Live Analytics Dashboard: Built with React.js and Chakra UI to visualize mentions processed, response rates, and activity feeds. • Self-Optimizing Polling: Adaptive logic manages Twitter’s strict rate limits while ensuring timely responses. • Error Handling & Logging: Robust system with clear logs for easy monitoring and debugging. • Tech Stack: React.js, Chakra UI, Node.js, Express, Twitter API, OpenRouter, TypeScript. • Challenges Solved: Twitter rate limits, React hook order issues, realistic demo generation. • Result: Delivered a scalable, production-ready AI-powered platform that reduces manual effort and boosts authentic interaction. Duration: seven to thirty days · Budget: $500 · Started 2025

I developed a full-stack educational e-commerce website for Homage Publishers, a company offering curriculum-aligned books. The goal was to build a responsive, AI-enhanced platform with complete admin control and secure payment functionality. • Built with MERN stack (MongoDB, Express.js, React, Node.js) • Admin dashboard for managing products, events, orders, and announcements • Integrated Resend.com for real-time email notifications on order approval • Implemented PayFast for secure, region-specific payment handling • Added AI-based features for smart product suggestions and admin tagging • Designed dynamic sections like Shop, Categories, New Arrivals, Events, and Announcements • Deployed and optimised for performance and scalability A key challenge was aligning real-time admin actions with automated email and payment processes. I solved this by building clean backend logic with async workflows and robust testing. The platform is now live and helping streamline educational book distribution through a user-friendly experience for both customers and admins. Duration: seven to thirty days · Budget: $500 · Started 2025

Northstay is a custom built vacation rental booking platform developed for a short-term rental company in Norway. The goal was to move beyond standard WordPress solutions and create a fast, scalable, and fully controllable system tailored to their business operations. The platform was built using a modern custom stack with a dynamic frontend, direct Hostaway integration for property sync, bookings, availability, and pricing, and a secure backend with an admin dashboard. Key features include direct bookings with Stripe payments, advanced property search and filtering, activities and shop sections, homeowner KPIs, multilingual support, and a mobile-first responsive design. One of the main challenges was handling real-time pricing, availability sync, and operational workflows without relying on heavy plugins. This was solved through direct API integrations and a custom admin system that allows the client to manage content, properties, bookings, and users independently without needing a developer. The result is a launch ready, high-performance booking platform designed for long term growth, scalability, and future AI enhancements. Duration: 1-3 months · Budget: $2000 · Started 2025

Prime World Markets is a custom built trading platform website developed for a fintech startup entering the online trading and investment space. The client wanted a sleek, modern, and highly interactive website inspired by industry leaders such as Deriv, Olymp Trade, and XM, while keeping full flexibility to integrate with their external trading technology provider. The project involved complete UI UX design, branding assets, and front-end development for all public-facing pages, including landing pages, product sections, onboarding flows, and mobile responsive layouts. Strong focus was placed on animations, smooth transitions, and performance optimisation to ensure a premium, professional feel across devices. I also handled API ready architecture, external platform redirections, CRM and broker platform integration points, SEO friendly structure, domain and DNS setup, deployment, and production launch. The website was built to meet compliance requirements, support future trading integrations, and scale as the business grows. The final result is a high performance, production ready fintech website with strong branding, smooth UX, and a scalable foundation for ongoing trading operations. Duration: 1-3 months · Budget: $3500 · Started 2025

Spreadfy.ai is an AI powered content generation platform designed to help users create branded text, images, and videos for social media and marketing. My role was as the lead full-stack developer, responsible for stabilising the existing product, migrating core infrastructure, fixing critical bugs, and adding new features. The project began with an unstable backend, broken billing flows, and incomplete AI integrations. I migrated the authentication system to Clerk, moved data to PostgreSQL, resolved Stripe and trial system issues, and rebuilt essential components to ensure reliable content generation and user flows. I also implemented key new features, including responsive landing pages, a share system, pricing tier consistency, enhanced UI/UX fixes, and social media template support. Throughout the project, I focused on quality, scalable design, and long-term maintainability to support public launch and future growth. Duration: 1-3 months · Budget: $1000 · Started 2025

QualiFy AI is an AI powered B2B lead intelligence platform designed to help sales and growth teams automatically score, qualify, and prioritise leads. The platform pulls leads from sources such as Apollo and UpLead, analyses them against an Ideal Customer Profile, and pushes high intent leads directly into connected CRM systems. My role involved building the full stack SaaS platform, including the AI scoring engine, lead dashboards, CRM-ready data output, and real-time syncing workflows. I implemented GPT powered lead analysis, intelligent tagging, role-based team workspaces, and subscription-based access with secure authentication. A key challenge was ensuring scoring accuracy while keeping the system fast and scalable for large lead volumes. This was handled through structured data pipelines, modular AI prompts, and optimised backend logic. The final result is a production ready SaaS product that helps B2B teams save time, improve conversion rates, and focus on closing high quality opportunities. Duration: 1-3 months · Budget: $3000 · Started 2025

Python Developer & Automation Specialist

Developed a Python application that automatically fetches daily and hourly weather data, generates detailed graphs and tables, and sends them via email with attachments. The solution uses OpenWeatherMap API, Matplotlib, and Seaborn for visualization, ensuring clear and professional weather reports delivered directly to your inbox.