Frontend
Responsive product interfaces focused on usability, performance and maintainability.
Yessine Eleuchi
Full-stack engineer focused on product engineering, backend systems and applied AI — from interface and architecture to data, automation and production.
Product · Systems · Intelligence
Selected work ↓02 / Capabilities
I work across interfaces, backend systems, applied AI and software delivery to build complete, maintainable products.
Responsive product interfaces focused on usability, performance and maintainability.
APIs, business logic, asynchronous workflows and reliable data systems.
AI capabilities integrated into practical product and business workflows.
Reproducible environments, automation and reliable software delivery.
03 / Selected Work
Full-Stack & Applied AI PixiEstim An intelligent software project estimation platform that transforms heterogeneous project documents into structured, AI-assisted software estimations through controlled LLM workflows, historical RAG context, deterministic business rules, and human validation.
Full-Stack Business Automation TejFlow A business automation platform designed for accountants to retrieve, classify, archive, and manage Tunisian TEJ certificates while simplifying reporting and repetitive administrative workflows.
Backend Engineering & Observability System Monitoring Application A monitoring platform built with FastAPI to track endpoint availability, application performance metrics, and threshold violations through a centralized backend service.
Computer Vision & Applied AI Automatic License Plate Recognition An AI-powered backend combining automatic license plate recognition, vehicle and violation management, and a RAG-based legal assistant for Tunisian traffic-law queries. 04 / Engineering
Scroll-driven system
01 / Idea
Before frameworks and architecture, there is a problem to understand, simplify and turn into a product direction.
02 / Product
Product engineering connects user experience, interaction and business logic into something people can actually use.
03 / Systems
APIs, databases, queues, workers and asynchronous workflows become the infrastructure behind the product.
04 / Intelligence
LLMs, RAG, document processing and computer vision only belong in the system when they create real product value.
05 / Ship
Designed, connected, tested and delivered as software that can run, evolve and be maintained.
Notes
Along with coding, I also like to write about life and technology. Here are some of my posts.
A practical comparison of the PDF extraction libraries I tested while building my estimation system, and how latency became the deciding factor.
A small post about how Zustand helps me manage shared state inside TejFlow without overcomplicating the React codebase.
A short personal note about how basic HCI and ergonomics principles ended up shaping the UI of TejFlow more than I expected.
06 / Contact