Building high-scale systems & AI applications.

Architecting production-grade Generative AI systems — multi-agent orchestration, image generation pipelines, and scalable backend infrastructure powering global fashion brands.

• Architected a dynamic multi-agent orchestration system in JavaScript — a runtime orchestrator that routes tasks across role-specialized agents (research, generate, validate) via a custom skills framework, powering multiple AI workflows across production systems.
• Designed a skills-based agent framework enabling modular capability assignment per agent, decoupling orchestration from execution — improving extensibility and enabling rapid deployment of new AI workflows without re-engineering core pipelines.
• Migrated the core image generation pipeline to a next-gen architecture, significantly reducing production failure rates and elevating performance across 500K+ monthly generations.
• Engineered an end-to-end AI-powered marketing outbound pipeline — ingesting CSV user lists, auto-scraping brand sites for product extraction, computing semantic brand-aesthetic embeddings for template matching, and triggering personalized image generation for hyper-personalized email campaigns at scale.
• Built brand & template personalization using cosine similarity, matching fashion brands to hyper-personalized templates aligned with their brand guidelines.
• Built a custom training pipeline with AI Toolkit and Flux Context Portrait models, producing 500+ fine-tuned fashion models for editorial generation at scale.
• Engineered Bulk Click Generation from scratch — the company’s second customer-facing feature launch — enabling scalable bulk editorial image generation and producing 500K+ images in production.
• Designed and maintained multiple AI image transformation pipelines (image-to-ghost, image-to-sketch, image-padding) powering editorial tooling that generates 250K images monthly across global fashion brands.
• Implemented interactive in-product guided tours and continuous UI/UX enhancements, improving platform stability, user activation, and retention.

Sole engineer for SyneHQ — an AI-native data analytics platform where teams query their data in natural language. Built the product end to end as a freelance engagement, from architecture to launch.

• Designed and built the entire platform single-handedly — frontend, backend, AI query layer and deployments — taking the product from zero to a launched, customer-facing platform in 4 months.
• Built the AI-assisted analytics core: natural-language querying over connected data sources with SQL/Python notebook workflows for analysts.
• Implemented multi-source data integrations and the connection layer powering centralized analysis across customer warehouses and databases.
• Worked directly with the founder, owning scope, technical decisions and delivery end to end — the engagement every freelancer wants to be trusted with.

Full-stack product and high-fidelity frontend work — story-driven landing experiences and scalable backend systems.

• Developed 15+ interactive landing pages in Next.js with advanced scroll animations using GSAP & Framer Motion — crafting smooth, story-driven experiences aligned with brand identity that increased engagement.
• Architected a scalable notification system for custom notifications over text / WhatsApp / email.
• Collaborated with 4+ developers to build an e-commerce platform capable of supporting 1M+ monthly visitors.

Full-stack development across mobile and web for internal tools and client applications.

• Developed a mobile application from scratch using React Native + TypeScript, resulting in a 20% increase in hiring efficiency in the third quarter.
• Designed and implemented the front end of the terminal application using React.js + TypeScript, boosting admin efficiency by 30%.
• Migrated the backend of the employee portal from PHP to Node.js.
• Developed APIs for the dispatch application using Node.js as the backend framework.