Full-Stack / AI Engineer · Algiers, Algeria
I build production web platforms, computer-vision pipelines, and LLM-integrated systems with TypeScript, Next.js, NestJS, FastAPI, and Python. I hold a Master's degree in Intelligent Computer Systems Engineering, focused on full-stack product engineering and applied AI.
6 selected projects across production platforms, AI research, client delivery, and agentic planning.
Multi-agent business-intelligence assistant for Djezzy's commercial department: natural-language questions over 13 months of daily commercial data, answered with narrative text and live charts through a durable, asynchronous agent pipeline.
Built role-based modules for a TPE operations platform managing bank requests, tickets, inventory, attachments, Excel workflows, and notifications using Next.js, NestJS, Prisma, PostgreSQL, JWT, CASL, and SSE.
Research pipeline for prostate whole-slide images using OpenSlide preprocessing, HistoEncoder/Virchow2 embeddings, artifact gating, ABMIL/TransMIL, QWK evaluation, and attention heatmaps.
Prototype smart aquaculture platform with fish counting, disease checks, biomass estimates, pond monitoring, alerts, reports, mobile flows, and an admin dashboard.
Trilingual, RTL-first commerce and CMS platform for an Algerian marketing agency: a public site, an admin backoffice, and a versioned form-schema engine, live on the client's own infrastructure.
Second-place Tattwir prototype for AI-assisted business planning: sales forecasts, production planning, inventory alerts, external context, and action recommendations.
A production platform, a research result, a model behind a product screen, and a pipeline that needed no model at all.
Arabic, French and English from one content tree, RTL as the default. Measured on the client's own VPS: 87ms to first byte, 1.02s to largest contentful paint.
How it was builtTransMIL over Virchow2 embeddings with artifact gating, evaluated by quadratic weighted kappa across six ISUP grades. End-of-study project, graded 17.75/20.
How it was builtA YOLOv8 counting and disease-detection pipeline surfaced as something a fish farmer reads on a phone: pond health, biomass estimates and alerts.
How it was built
Otsu thresholding and morphological cleanup separating overlapping coins on a noisy board — the cheapest correct answer rather than the fashionable one.
How it was builtProfessional experience across full-stack delivery, computer vision, and applied AI.
Led a 4-person engineering team building the Djezzy Commercial Copilot, a multi-agent business-intelligence assistant for the commercial department, during a 6-week internship sprint.
Freelance software engineer building web platforms for nightbyte, managing delivery end to end from architecture through deployment.
Built internal service-management software and cloud tooling for TPE operations.
Built computer-vision and product features for internal CRM/ERP and AI assistant work.
Built an LLM virtual-assistant prototype with text and voice interaction modes.
Technologies grouped by current depth, so the strongest production stack is clear at a glance.
Open to opportunities, collaborations, and technical conversations about engineering and research.