Latest Projects
Explore My Most Recent Creations
Independent R&D Rayni — AI Research Chatbot
Click here to view the live app
Click here to view the code
A next-generation chat interface built for AI-assisted document reasoning and contextual knowledge retrieval.
The system is architected around a modular LLM orchestration layer that supports retrieval-augmented generation (RAG),
adaptive context-window compression, and semantic reranking of document snippets in real time.
Users can upload or import PDFs from Google Drive and engage in intelligent dialogue grounded in document content,
with streaming token-level responses, citation-aware memory, and a feedback-reinforcement loop
that refines system prompts dynamically based on prior conversation state.
Implemented with React (Vite), TypeScript, Tailwind CSS, React-PDF, and a Node/Express micro-API.
Deployed on Vercel with Google OAuth2 and a configurable adapter layer for multi-model support
(OpenAI, Anthropic, Bedrock).
Given my background in cognitive neuroscience research, I couldn’t help but imagine how invaluable this kind of
lab-friendly AI chatbot would’ve been—especially during those late nights labeling EEG data and arguing with MATLAB.
this would’ve been the only lab partner who never asked for coffee breaks.
Client Engagement Campers Coffee — Veteran-Owned Shopify Platform
Click here to view the live site
Click here to view the website review from my client, Jonathan
A full-stack e-commerce platform built for Campers Coffee, a veteran-owned business that funds City Campers — a nonprofit dedicated to providing direct support to the homeless.
This project began as a complete migration from a rigid third-party builder to a fully custom Shopify deployment, designed and engineered from the ground up.
Beyond the storefront, I implemented secure employee authentication, custom payroll management tools, and dynamic collection pages driven by real-time product data.
The backend integrates lightweight AI-powered content optimization to generate marketing copy variations and sentiment-aware product descriptions based on user engagement metrics.
I also developed automated inventory and pricing sync scripts that scrape partner distributors’ data, normalize it, and feed updates directly into Shopify via a private API, ensuring zero downtime and fresh listings daily.
Built with Shopify Liquid, JavaScript, Node.js, REST APIs, and custom SCSS architecture.
Integrated serverless cron jobs for analytics and AI-driven email targeting to highlight seasonal blends and community stories.
As someone who’s passionate about human-centered design, this project was more than code — it was building technology that amplifies a mission.
Plus, it’s the only coffee site that can legitimately say every purchase helps someone get back on their feet — and that’s a backend worth maintaining.
Independent R&D AI Web Scraper — Autonomous Web Intelligence Engine
Click here to view the live app
Click here to view the code
AI Web Scraper is an intelligent, end-to-end content analysis platform designed to autonomously extract and synthesize information from any website in real time.
The system leverages natural language processing (NLP) and semantic understanding to transform raw web data into concise, human-readable summaries — complete with topic detection, sentiment analysis, and keyword clustering.
Engineered with a focus on scalability and automation, the platform integrates a Node/Express API layer for distributed scraping tasks and a React-based analytical dashboard for visualization and insight delivery.
The underlying AI models interpret contextual relationships across pages, detect tone, and classify high-value information streams — enabling users to instantly understand the essence of any domain or article.
The application supports asynchronous operations, parallel parsing, and adaptive throttling, ensuring efficient data extraction at scale while maintaining responsiveness and reliability.
Deployed on Render and Vercel, it represents a fully containerized, production-grade microservice architecture that blends AI reasoning with modern web performance engineering.
This project demonstrates the fusion of intelligent automation and user-centric design — a web tool that not only gathers data, but genuinely understands it.
Independent R&D Crypto Order Book Aggregator — AI-Enhanced Market Dashboard
Click here to view the live app
Click here to view the code
Crypto Order Book Aggregator is a real-time Python + Flask web application that unifies live BTC/USD order-book data from
Coinbase, Gemini, and Kraken to display the most accurate cross-exchange liquidity picture available.
It dynamically calculates the cost to buy or sell Bitcoin at market using a liquidity-weighted algorithm that simulates
institutional trade execution.
The application features a modern dark UI, an embedded live BTC/USD chart, and a local AI-style market insight
module that analyzes spread tightness to summarize real-time market conditions.
Designed for simplicity and transparency, it’s deployed entirely free on Render with a lightweight architecture that
proves how far thoughtful engineering can go without heavy infrastructure.
This project combines financial data engineering, clean design, and intelligent UX — transforming complex crypto market
data into something instantly understandable and visually compelling.
Client Engagement Doubletake — Intelligent Shopify Experience
Click here to view the live site
Doubletake was founded by two entrepreneurs with a shared mission: to reinvent the sports-bag industry with equal parts functionality and modern design.
I was brought on as a freelance engineer to architect and develop their new e-commerce platform from the ground up — transforming early design sketches into a fully realized digital storefront.
The build leveraged Shopify Liquid, JavaScript, and custom SCSS frameworks to create a modular, high-performance theme with dynamic merchandising sections and predictive product surfacing.
Using lightweight ML-driven analytics hooks, the site adjusts featured products and layout hierarchy based on engagement data, inventory velocity, and real-time conversion feedback.
To streamline operations, I implemented automated SEO optimization and AI-generated content modules that adapt tone and descriptions according to user demographics — enabling Doubletake’s founders to maintain a premium brand voice without manual copywriting overhead.
Built with a focus on conversion intelligence, accessibility-first design, and responsive micro-interactions, the site embodies the brand’s mantra: “Less Talk. More Game.”
Every line of Liquid in this build was designed to perform double-duty — balancing elegance and analytics.
It’s what happens when custom development meets adaptive intelligence — a digital court where data quietly serves style.
Client Engagement M.O.T — African Minimalist Fashion Platform
Click here to view the live site
M.O.T (Mode of Thought) is a fashion label redefining African Minimalist Fashion through color, balance, and versatility.
I partnered with the founders to translate their design ethos into a digital experience — crafting the brand’s website from the ground up with a focus on performance, storytelling, and intelligent automation.
Built on Shopify Liquid with custom JavaScript and SCSS modules, the platform integrates a suite of AI-assisted merchandising tools and dynamic product curation.
Using a lightweight machine-learning layer, the system identifies seasonal trends and customer preferences to automatically reorder featured products and surface color palettes aligned with real-time user behavior.
To maintain consistency with M.O.T’s design philosophy, I implemented adaptive layout algorithms that intelligently adjust visual hierarchy across devices — balancing form and function while preserving the brand’s minimalist aesthetic.
Additional features include AI-optimized SEO tagging, automated lookbook generation, and multi-region language detection for international audiences.
This project merged fashion, culture, and technology — where neural networks meet neutral tones.
It’s an example of how data-driven design can feel organic, beautiful, and unmistakably human.
Client Engagement Lamazi Fabrics — Intelligent E-Commerce for Sustainable Design
Click here to view the live site
Lamazi Fabrics is a family-run business built by James and Liana — passionate creators curating premium, sustainably sourced sewing supplies for a global community of makers.
I partnered with them to design and engineer a custom Shopify experience from the ground up that captures the warmth of a handmade brand with the intelligence of a modern digital storefront.
Developed with Shopify Liquid, JavaScript, and SCSS architecture, the site integrates AI-assisted product categorization and machine-learning powered search ranking to help customers discover fabrics through color similarity, texture, and sustainability tags.
Behind the scenes, automated inventory monitoring scripts and predictive restock notifications ensure customers never miss high-demand materials.
I implemented AI-driven merchandising logic that adapts featured collections based on seasonal interest, engagement heatmaps, and user browsing behavior — all while preserving the brand’s artisan storytelling.
The design blends minimalist UI principles with data-informed layout adjustments to optimize conversions without losing aesthetic balance.
Working with Lamazi was about weaving technology into craft — an e-commerce ecosystem that feels personal, sustainable, and quietly intelligent.
Every line of Liquid in this build has its own thread of logic — stitched with precision and purpose.
Client Engagement Chama Valley Meat Co. — Smart Local Commerce Platform
Click here to view the live site
Chama Valley Meat Company is a family-owned, community-first business serving high-quality, locally sourced meats to their New Mexico region.
I collaborated with another developer to rebuild their online presence from the ground up — delivering a platform that fuses artisanal authenticity with intelligent digital infrastructure.
Built on Shopify Liquid with custom JavaScript and modular SCSS architecture, the site integrates AI-powered inventory forecasting and demand prediction models to anticipate sales surges and optimize product availability.
I implemented dynamic recipe pairing recommendations using a lightweight semantic-matching model that suggests complementary products (e.g., rubs, sauces, and sides) based on natural-language ingredient data.
The storefront also features AI-driven email personalization and automated content scheduling — ensuring local customers always see the freshest offers, promotions, and seasonal cuts.
Every layout was hand-coded with performance in mind, maintaining sub-second load times and strong SEO signals across regions.
This build was about more than e-commerce — it was about using intelligent automation to empower a local business that feeds its community.
I like to think of it as farm-to-table, reimagined as code-to-commerce.