Adrian stack's Work | ContraWork by Adrian stackAdrian stack
AI engineer and website developer
Architected and developed a full-stack AI SaaS application using the MERN stack to automate high-converting thumbnail ideation and creative workflows for content creators.
Integrated Google Gemini API to build an intelligent prompt pipeline that translates video topics and metadata into engaging visual concepts.
Engineered a secure data persistence layer using MongoDB to efficiently store, retrieve, and manage multi-tenant user history and generated assets.
Implemented robust session handling and protected React routes via secure user authentication to ensure isolated and private creator dashboards OmniScribe is a production-grade, full-stack AI SaaS workspace designed to bundle diverse content generation and document productivity utilities into a single application. Built with React.js, Node.js, and Express.js, the platform utilizes serverless PostgreSQL hosted on Neon for reliable, scalable data persistence. OmniScribe features a modern user interface that bridges multiple downstream AI and Computer Vision APIs—including text-to-image generation, automated image editing, and semantic resume analysis—while relying on Clerk for enterprise-grade authentication and user access management. Engineered a scalable multi-vendor e-commerce marketplace using Next.js and PostgreSQL, implementing multi-tenant architecture and dynamic role-based access for admins, vendors, and customers.
Integrated Stripe API to handle split-payment processing, checkout workflows, and multi-tier premium seller subscriptions.
Designed an asynchronous background job pipeline using Inngest to manage scheduled tasks, webhook events, and marketplace automated workflows.
Optimized asset delivery and state management by leveraging ImageKit for cloud image compression, Clerk for session authentication, and Vercel for high-availability serverless deployment. Architected and deployed a production-ready AI SaaS application using the PERN stack (React, Node.js, Express, PostgreSQL) to automate high-converting marketing content generation.
Integrated Google Gemini API to architect a prompt-engineered pipeline delivering real-time, context-aware advertising copy and creative frameworks.
Engineered secure user workflows by implementing Clerk for multi-tenant authentication and consumption-based billing infrastructure.
Optimized data reliability and uptime using Neon serverless PostgreSQL for data persistence, Sentry for real-time error tracking, and a Hostinger VPS for scalable cloud deployment.