Marceli Borowczak

Built an internal lead-delivery platform that synchronized user-submitted web domain sources with multiple external providers (Iterable, SFMC, Everflow).

• Designed and developed the full frontend in Next.js, using modern React standards, server components, app router, and type-safe API integration.
• Architected a high-performance backend in NestJS (Node.js) capable of handling up to 10,000+ requests per second, featuring scalable routing, queueing, rate control, and parallel ingestion pipelines.
• Implemented robust domain-parsing logic, data validation, provider routing rules, and automated retry mechanisms.
• Delivered a reliable internal product that reduced operational overhead and streamlined outbound lead distribution across multiple marketing ecosystems.

• Developed and maintained advanced UI systems for the Viberp GTA V roleplay server running on alt:V, focusing on performance, clarity, and modularity.
• Built front-end interfaces using Vite, TypeScript, TanStack Query, and TanStack Router, ensuring fast load times and smooth in-game interactions.
• Implemented real-time UI flows for gameplay systems such as inventories, shops, notifications, character panels, job mechanics, and administrative tools.
• Created reusable, type-safe UI component libraries optimized for alt:V WebView rendering.
• Integrated UI with backend gameplay logic through structured event bridges, RPC calls, and server→client sync mechanisms.
• Improved architectural consistency by introducing state management patterns, request caching strategies, and modular routing.
• Optimized build performance, reduced bundle size, and enhanced stability across different client hardware setups.

• Designed and built the complete application architecture: Next.js 15 frontend, NestJS backend, PostgreSQL database, Redis caching layer, and distributed job scheduling (CRON).
• Implemented core planning logic, multi-step AI chat workflows, itinerary generation, user preference engine, and data ingestion pipelines.
• Developed a custom AI model for itinerary generation and destination ranking using LoRA/QLoRA fine-tuning on Mistral-7B, including dataset creation, embeddings, and inference optimization.
• Integrated Amadeus API for flights and hotels; built internal services for flight search, pricing, caching, normalization, and fallback logic.
• Created scalable backend modules for authentication, user profiles, subscriptions, points system, activity logs, token usage tracking, and vector search.
• Optimized backend performance with efficient query patterns, request batching, background jobs, and structured logging across services.
• Implemented the full Next.js frontend with server components, SSR, type-safe API routes, global state management (Zustand), and real-time itinerary editing.
• Deployed infrastructure across Vercel and DigitalOcean, configured CI/CD pipelines, environment isolation, monitoring, and error tracking.

• Developed advanced game UI interfaces for a large-scale GTA V roleplay server running on alt:V.
• Built modular, high-performance frontends using Vite, TypeScript, TanStack Query, and TanStack Router.
• Implemented client-side state management, server→client communication handlers, and optimized data-fetching flows for low-latency in-game interactions.
• Designed reusable UI components and layouts for inventory systems, dialogs, character menus, vehicle panels, job systems, and other in-game mechanics.
• Collaborated closely with backend/game-logic developers to integrate UI events, RPC calls, and real-time synchronization between the game client and server.
• Ensured fast builds, code-splitting, and seamless integration with the alt:V WebView environment.
• Refactored legacy UI modules into a modern, type-safe architecture with improved performance and maintainability.