AI Development
I build AI-powered features and LLM-driven workflows — from automation and data pipelines to intelligent tooling that turns business requirements into production-ready systems.
Hello, I'm
LeshanPatterson
Senior Full-Stack Developer specializing in AI development, full-stack engineering, and cloud engineering.
0Years Experience
0+Happy Clients
0+Completed Projects
About Me
Full Stack Development
I deliver scalable, production-grade systems end to end — React and Next.js front-ends, Node, PHP, and .NET back-ends, and the REST APIs that connect them, built for performance and maintainability.
Cloud Engineering
I architect and deploy on AWS and Google Cloud with CDN-backed, automated CI/CD pipelines — optimizing reliability, scalability, and delivery speed across environments.
Work Experience
HiddenGem Media
March 2025 – PresentFront-End Developer
- Lead end-to-end development of complex WordPress systems, including custom plugins, APIs, and component-based templates using PHP and JavaScript.
- Architect data-driven features and third-party integrations (analytics, booking, CRM) to improve performance, reliability, and conversion tracking.
- Partnered directly with executive stakeholders to translate business objectives into scalable technical solutions.
- Design and maintain scalable WordPress architectures optimized for performance, security, and long-term maintainability.
- Diagnose and resolve complex production issues across hosting, CDN, and application layers.
MAPL TECH
October 2024 – PresentCo-founder & Tech Lead
- Co-founded and scaled MAPL TECH, leading technical strategy and delivery across multiple markets, surpassing $10K in revenue within three months and establishing operational branches in Jamaica and Nigeria.
- Build AI-powered workflows and automation systems, integrating data pipelines, analytics, and intelligent tooling to improve efficiency, decision-making, and platform scalability.
- Lead end-to-end development of production-grade systems, including custom WordPress plugins, APIs, and reusable component frameworks, partnering directly with executive stakeholders.
spark* advocacy
April 2022 – Oct 2024Full Stack Developer
- Developed high-performance websites using WordPress, PHP, JavaScript, CSS, HTML, and Twig.
- Led end-to-end website development projects, managing lifecycles from concept to launch.
- Optimized CMS and static websites for performance, accessibility (WCAG AAA), and SEO.
CTRLALTDEL.IO
Jan 2021 – Nov 2021Front End Developer
- Optimized WordPress sites for performance, SEO, and conversions, achieving a 20% increase in performance.
- Integrated complex APIs to enhance website functionality and user engagement.
- Designed responsive, user-friendly websites using Figma, Adobe XD, and modern web standards.
Education
Google UX Design
Dec 2022 – Jun 2023Professional Certificate
- Create designs on paper and in digital design tools like Figma and Adobe XD
Lighthouse Labs
Nov 2021 – Feb 2022Diploma In Web Development
- JavaScript/Node.js
- Software Architecture
- React Js
- Ruby on Rails
Udacity
July 2021 – Oct 2021AWS Machine Learning Scholarship
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
University of Waterloo
2013 – 2017Computer Science
Featured Projects
AI can write the code. The harder problem is building systems you can trust to write it.
Over the last several months, I have been developing my agentic engineering practice by building and testing production systems where AI agents handle much of the implementation, while I focus on architecture, verification, guardrails, and the decisions around them.
MAPL Tours
Putting Sentinel to the Test in Production
A real production environment for testing how far agentic engineering can go beyond prototypes: video-first discovery, live quoting, payments, dispatch, and lifecycle email. This is not a demo. It processes real payments for rides operated by a licensed transportation provider in Jamaica, so the engineering has to account for real-world failure modes.
The system includes
- A project specification every agent reads before making changes, plus persistent decision memory
- Nearly 2,000 tests running through a CI/CD pipeline before changes reach production
- Adversarial review, where specialist agents critique changes and a skeptic agent tries to disprove the findings
- Hard financial guardrails including server-side pricing, idempotent payments, and isolated test traffic
- Agent-native interfaces, including MCP and WebMCP tools that allow AI assistants to query real fares and hand users a ready-to-book link
~50%Roughly 50% of customers now arrive through ChatGPT
Sentinel
Design Intelligence as Infrastructure
Models can generate UI that works. Getting them to consistently apply good design thinking before they start building is a different challenge. Sentinel is a design agent exposed through MCP that a model can call before it builds anything visual.
It brings design thinking in through
- Problem framing and design-thinking guidance
- User intent, needs, and journey analysis
- Information architecture and content hierarchy
- Visual hierarchy, interaction, and layout patterns
- Contextual design critique and iteration guidance
Shared memory
Every design Sentinel analyzes and every score it produces feeds shared memory, so each iteration starts with more context than the last.
589258 to 92 score on one production page after rebuilding against its recommendations
The biggest lesson
Generation is only one part. Fundamentals and engineering best practices still matter.
AI can accelerate implementation, but it does not remove the need for architecture, testing, security, CI/CD, observability, documentation, design thinking, and disciplined engineering practices. If anything, those fundamentals become more important when agents are producing more of the code.