Full Stack Intern @ Mentozy | Backend Dev @ Hoollow | Web Dev @ Glorify | Lead Dev @ Grindly | Founder & CEO @ Meetify

github

medium

devto

dribbble

leetcode

projects

Raymond Oyondi  • Peerlist

Linkify is a polished, highly responsive Link-in-Bio and multi-channel social media management platform designed for modern creators and teams. Built to bridge the gap between static landing pages and complex operations, it gives users complete control over their digital presence.The platform features a fully customizable frontend layout alongside advanced administrative controls, making it easy to create cohesive, secure, and data-driven hub spots for all social profiles.🚀 Key Features<ul><li>Custom Theme Presets — Personalize your landing configurations with elegant, responsive design templates.</li><li>Real-Time Clickstream Analytics — Monitor live user interaction and track event metrics as they happen.</li><li>Dynamic Role-Based Permissions — Protect infrastructure routes using granular, custom security access rules.</li><li>Automated PDF Reports — Generate performant, publication-ready metric reports instantly for client handoffs.</li><li>Hot Module Replacement (HMR) — Enjoy hyper-fast development speeds driven by a modern compiler environment.</li></ul>

Linkify - Link-in-bio tool

Project OverviewRestroby is an advanced, production-ready Artificial Intelligence (AI) platform engineered specifically for the Food and Beverage (F&amp;B) sector. The application serves as a comprehensive operational operating system that leverages autonomous, multi-agent AI workflows to eliminate manual data entry, streamline complex back-of-house logistics, track kitchen waste, and optimize supply-chain ordering patterns. Built using a hyper-modern and type-safe architecture, the platform features a reactive frontend built with Vite and TypeScript, backed by an enterprise-tier Node.js micro-backend and micro-applet structures managed via Google Firebase.The core philosophy behind Restroby is to empower restaurant owners, kitchen managers, and enterprise hospitality operators with the tools necessary to combat shrinking profit margins, volatile supply chain pricing, and food waste. By implementing modern event-driven AI agents, Restroby automates the heavy lifting of back-of-house management, allowing restaurant staff to focus entirely on culinary execution and hospitality delivery.Core Architectural Deep DiveRestroby is organized as a unified, highly optimized hybrid repository containing both structural client-facing logic and a centralized backend server orchestration engine. The architecture prioritizes separation of concerns, explicit schema declaration, and secure multi-tenant data boundaries:<ul><li>Frontend Tooling &amp; Core App Layer (/src &amp; index.html): Built on top of Vite, the application utilizes a blazing-fast Hot Module Replacement (HMR) development server and generates optimized, chunk-split static production assets via Rollup. Strict TypeScript rules ensure compiled code is resilient against runtime crashes. </li><li>Backend Server Core (server.ts): A lightweight, highly performant Node.js server instance handles external API connectivity, processes compute-intensive AI operations, coordinates multi-agent task loops, and securely communicates with third-party supplier platforms.</li><li>Firebase Applet System Architecture (firebase-applet-config.json &amp; firebase-blueprint.json): Restroby utilizes a unique decoupled applet system schema. This structure sandboxes independent tenant restaurant branches, allowing them to spin up localized modules (e.g., specific POS integrations or specialized vendor workflows) without altering the global codebase core.</li><li>Granular Multi-Tenant Security (firestore.rules): Multi-tenancy isolation is enforced at the database layer. Custom Firestore security rules ensure that restaurant chains, local store operators, kitchen captains, and third-party vendors can only query and modify data partitions explicit to their verified permission groups.</li></ul>Key Innovations &amp; Features🤖 Autonomous AI Agent WorkflowsAt the heart of Restroby is an event-driven AI agent runtime. Instead of relying on rigid, hardcoded conditional automation rules, the platform deploys autonomous agents that analyze operational telemetry. These agents can dynamically identify anomalies—such as an unlogged stock drop, an unexpected price spike from a vendor, or a persistent delay in supplier deliveries—and automatically initiate background tasks like sending alerts or drafting communications.📦 Predictive &amp; Smart Inventory ControlRestroby changes the inventory management paradigm from a reactive chore to a predictive automated workflow. By analyzing historical consumption patterns, seasonal menu changes, local weather events, and historic supplier lead times, the platform maps out highly precise par levels. It tracks ingredients in real time and automatically creates pending purchase orders before a kitchen experiences a service-disrupting stockout. 🍏 Kitchen Waste Tracking &amp; AI AnalyticsFood waste directly degrades restaurant profitability. Restroby includes an interface to easily log, bucket, and analyze daily kitchen scrap and spoilage data. The AI pattern recognition engine matches waste trends against prep lists and menu popularity. For example, if the platform flags that a prep cook is consistently over-preparing fresh ingredients on Tuesday mornings, it will adjust the automated prep schedules to match historical Tuesday demand.🧾 Automated Optical Invoice Parsing &amp; ReconciliationManually keying in long, line-item supplier invoices is both error-prone and time-consuming. Restroby automates this ingestion process through an intelligent document parsing pipeline. Users drop physical invoice images or supplier PDFs into the interface; the platform extracts item codes, quantities, and line item totals, automatically cross-checks them against the initial purchase order, updates inventory quantities, and logs financial changes.

Restroby - AI-powered F&B platform

🏗️ Grindly: The End-to-End Engineering EcosystemThe VisionMost coding platforms stop at syntax. They teach you how to write a for loop or fetch an API, but they often leave a massive gap between "writing code" and "architecting systems." Grindly was built to bridge that gap.It is an immersive, all-in-one engineering ecosystem designed to move beyond simple tutorials. Grindly plunges users into the full Software Development Lifecycle (SDLC), providing the tools and environment needed to think like a Senior Engineer.🚀 Core Features1. Interactive Multi-file IDEAt the heart of Grindly is a robust, browser-based coding environment powered by the Monaco Editor (the engine behind VS Code). This isn't just a single-file playground; it supports multi-file project structures, allowing users to practice modular code organization and real-world project management. 2. Architecture CanvasEngineering is as much about drawing as it is about typing. Grindly features a drag-and-drop Architecture Canvas where users can design complex system architectures. Visualize how services interact, map out databases, and plan your infrastructure before a single line of code is written. 3. AI-Driven Architect ReviewerGet instant feedback from an "AI Senior Lead." The Architect Reviewer provides automated critiques of both your code and your system designs. It ensures your work aligns with industry best practices, flagging potential bottlenecks or anti-patterns in real-time.4. SDLC ImmersionGrindly focuses on the "how" and "why" of software. The environment emphasizes:<ul><li>Test-Driven Development (TDD): Building reliable systems through testing.</li><li>SOLID Principles: Enforcing object-oriented design excellence.</li><li>Scalable Architecture: Planning for growth using proven system design patterns.</li></ul>

Grindly - End-to-end engineering ecosystem

Meetify is a cutting-edge video conferencing application designed for modern enterprises. Built with Next.js 16, TypeScript, and Tailwind CSS, it offers a robust platform for real-time communication. Leveraging the Stream Video SDK, Meetify ensures scalable, low-latency, and high-quality video interactions. Key features include secure authentication through Clerk, instant and scheduled meetings, and comprehensive meeting recordings accessible from a personal dashboard. Users can enjoy real-time interaction with features like screen sharing, participant lists, and integrated chat. The application also provides a personal room with a dedicated, permanent link for recurring meetings. Its responsive design is optimized for desktop, tablet, and mobile viewing, making it versatile for various user needs. Meetify is deployed on Vercel, ensuring smooth and efficient performance.

Meetify - A modern video conferencing app

Shine is a high-performance e-commerce shopping cart application designed to provide a seamless user experience. Built with React and Redux Toolkit, it allows users to browse products, manage their cart in real-time, and calculate costs dynamically. Key features include:<ul><li>Real-time Cart Management: Add or remove products with instant UI updates.</li><li>Smart Quantity Adjustments: Easily increase or decrease item counts directly within the cart.</li><li>Dynamic Totals: Automatic calculation of subtotal, taxes, and grand totals using Redux selectors.</li><li>Responsive Design: Fully optimized for desktop, tablet, and mobile viewing.</li><li>State Persistence: Built with Redux Toolkit for efficient, predictable state transitions.</li></ul>The tech stack includes React.js for the frontend, Redux Toolkit for state management, and styling with CSS3 and Tailwind CSS. Icons are provided by Lucide-React and FontAwesome. Shine is open source, allowing developers to contribute and enhance its functionality.

Shine - E-commerce Shopping Cart

ResumeAI is an innovative full-stack web application designed to revolutionize the way job seekers create resumes. By leveraging the power of AI, this platform transforms raw career data into professional, ATS-friendly resumes in minutes. Key Features include AI-powered content generation that crafts compelling bullet points and professional summaries, ensuring your resume passes through Applicant Tracking Systems with built-in keyword analysis. The platform offers a real-time live preview, allowing users to see changes instantly with a side-by-side editing interface. Secure Authentication is provided through modern auth providers, ensuring user data protection. Users can choose from a variety of customizable templates designed to be readable by both humans and machines. Additionally, ResumeAI supports high-quality PDF export, making your resume ready for job applications. Tech Stack includes React, Next.js, Tailwind CSS for the frontend, and Node.js with Express for the backend. AI integration is achieved through Google Gemini API and OpenAI API, with PostgreSQL and MongoDB as database options. Authentication is managed by Clerk and NextAuth.

ResumeAI - AI Resume Builder

PromptStore is a cutting-edge AI Prompt Marketplace that empowers creators to monetize their prompt engineering skills while providing users with a curated library of high-quality prompts for models like GPT-4, Midjourney, and Claude. Built with a serverless-first architecture, PromptStore ensures scalability, speed, and cost-efficiency. The platform is designed using a modern tech stack including Next.js for the frontend, Node.js and TypeScript for serverless backend functions, and Prisma ORM with a scalable PostgreSQL database. Authentication is managed through NextAuth.js or Clerk, and payments are seamlessly handled with Stripe Integration. Deployment is managed via Vercel for high-availability hosting. Key features include browsing and filtering prompts by category or AI model, a user dashboard for sellers to track sales and manage listings, secure checkout with instant access to purchased prompts, and a responsive design optimized for all devices. The serverless approach allows the application to scale to zero when not in use, significantly reducing operational costs while maintaining the ability to handle traffic spikes.

PromptStore - AI Prompt Marketplace

CampusSwap is a comprehensive full-stack marketplace platform tailored for students, facilitating the buying, selling, and trading of items and services within campus communities. Built on the robust MERN stack (MongoDB, Express, React, Node.js) and enhanced with PostgreSQL for handling relational data, this application ensures a seamless and secure user experience. Key features include:<ul><li>Secure Authentication: Utilizes JWT for user login and registration, with student email verification for added security.</li><li>Hybrid Database Architecture: Combines the flexibility of MongoDB for real-time data with the reliability of PostgreSQL for structured transactions.</li><li>Real-Time Messaging: Enables instant communication between buyers and sellers through Socket.io.</li><li>Robust Search &amp; Filters: Allows users to filter listings by category, price range, or item condition.</li><li>Dynamic Listings: Users can easily upload, edit, and manage their items, complete with image support.</li></ul>The application is designed with a modular architecture to ensure scalability, featuring a responsive React SPA client, a Node/Express API server, and a data layer that supports both ACID compliance and horizontal scaling.

CampusSwap - Student Marketplace

On-Orbit Collision Predictor is a comprehensive MATLAB-based tool designed for satellite conjunction analysis and space debris collision risk assessment. As Earth's orbit becomes increasingly congested, the risk of catastrophic collisions grows exponentially. This project provides a robust framework to predict future positions of space objects using TLE (Two-Line Element) data, detect "Close Approach" events within a defined spatial volume, and quantify the probability of collision using established statistical models.<h3>Key Features</h3><ul><li>SGP4/SDP4 Propagation: High-fidelity propagation of TLE data sets to account for atmospheric drag, Earth oblateness, and solar/lunar perturbations.</li><li>TCA Determination: Accurate calculation of the Time of Closest Approach using iterative minimum-distance algorithms.</li><li>Probability Analysis: Implementation of Foster’s and Chan’s algorithms for collision probability estimation.</li><li>Covariance Transformation: Mapping of positional uncertainty ellipsoids into the encounter frame.</li><li>Visualizations: 3D orbital trajectory plotting and relative distance-over-time graphs.</li></ul>This tool is essential for ensuring satellite safety and managing space debris, providing users with the ability to propagate orbits, identify conjunctions, and assess collision risks effectively.

On-Orbit Collision Predictor

HoopsForecast is an advanced machine learning project designed to forecast player statistics for upcoming NBA seasons. This project leverages a comprehensive data science pipeline, integrating automated data scraping, feature engineering, and sophisticated regression modeling to provide accurate predictions. The core functionality is powered by an interactive web dashboard application, allowing users to visualize performance predictions and explore various 'what-if' scenarios.<h3>Key Features</h3><ul><li>Automated Data Pipeline: Efficiently scrapes and cleans player-per-game stats, advanced metrics, and team standings.</li><li>Feature Engineering: Calculates rolling averages, career trajectories, and age-based performance curves to enhance prediction accuracy.</li><li>Multi-Model Comparison: Utilizes Linear Regression, Random Forest, and Gradient Boosting (XGBoost) models to evaluate and improve prediction outcomes.</li><li>Web Dashboard: Offers an interactive interface for users to visualize individual player projections and analyze potential outcomes.</li></ul>The project is built using a robust tech stack including Python, Pandas, NumPy, and Scikit-Learn, among others. It is open source and available under the MIT License, encouraging contributions and collaboration from the community.

HoopsForecast - NBA Performance Predictor

Stocklytics offers a comprehensive solution for real-time stock analysis, leveraging advanced technical indicators, price forecasting, and news sentiment analysis. This innovative platform is designed for investors and traders who seek to make informed decisions based on data-driven insights. Key features include the integration of technical indicators that provide a detailed view of market trends, enabling users to anticipate price movements effectively. Additionally, the platform's price forecasting capabilities are powered by sophisticated algorithms that predict future stock prices with high accuracy. <ul><li>Real-time stock analysis</li><li>Advanced technical indicators</li><li>Price forecasting</li><li>News sentiment analysis</li></ul> By analyzing news sentiment, Stocklytics helps users understand market sentiment and its potential impact on stock prices. This feature is particularly valuable for gauging the mood of the market and making timely investment decisions. Whether you are a seasoned trader or a novice investor, Stocketics provides the tools and insights needed to navigate the complexities of the stock market with confidence.