Open to opportunities · Tunis, Tunisia

Youssef Jmel

Final-year CS student building AI-powered systems — from intelligent IT support platforms to data pipelines and full-stack products. I ship end-to-end: architecture, AI pipeline, polished UI.

About

Who I am

I'm a final-year Computer Science student at the Faculty of Sciences of Tunis, specialising in AI, data engineering, and full-stack development. I enjoy building systems that actually solve problems — from a smart IT helpdesk that reasons over historical tickets to real-time BI dashboards that replace spreadsheet chaos.

My flagship project is an AI-powered ITSM platform— a full production-grade system I've been engineering for my Cycle Ingénieur PFE at Teamwill. It brings LLM-driven ticket classification, intelligent chat guidance, proactive SLA monitoring, and Redis-cached recommendation pipelines to IT support teams.

I work efficiently with AI coding agents like Claude Code — using them not just to write code faster, but to reason about architecture, navigate large codebases, and maintain quality across a complex multi-service system.

Outside engineering, I managed media production for the FST Free Software Club and co-organised two editions of the GENESIS Competitive Programming Contest.

6+
Projects shipped
2
Internships
2026
Graduation

AI-First Engineering

Built production AI pipelines using Claude, NLP, and vector embeddings — including an LLM-powered ITSM copilot with multi-intent chat, proactive SLA monitoring, and auto-classification.

Data & BI

Designed end-to-end ETL pipelines on Azure + Microsoft Fabric, star-schema data warehouses, and real-time Power BI dashboards with custom DAX measures.

Full-Stack Products

Shipped real client projects — e-commerce platforms, CTF competition systems, and enterprise ITSM tools — covering backend APIs, databases, and polished React frontends.

AI-Assisted Development

Proficient with AI coding agents — I routinely use Claude Code to accelerate architecture decisions, scaffold features, and maintain large codebases with precision and speed.

Experience

Where I've worked

AI Engineering Intern — PFE

Active

Teamwill · Tunis

Feb 2026 – Present
  • Architecting and building a production-grade AI-powered ITSM platform as my 5-month final engineering project (Cycle Ingénieur).
  • Designed an LLM-driven AI copilot with multi-intent chat, contextual cause analysis, proactive SLA monitoring (75% threshold), and auto-classification of IT support tickets.
  • Implemented a two-level Redis caching layer (LRU + 24h Redis TTL) on embeddings and all major API endpoints — reducing repeated LLM calls significantly.
  • Built a full Alembic migration chain (36 versions), a vector-similarity duplicate detector (threshold 0.72), and a recommendation engine with confidence scoring and feedback loops.
FastAPINext.jsClaude AIPostgreSQLRedisNLPAlembicDocker

Data Engineering Intern

Teamwill · Tunis

Jul – Aug 2025
  • Designed and delivered end-to-end data pipelines on Azure and Microsoft Fabric, eliminating ad-hoc spreadsheet reporting for the business team.
  • Built interactive Power BI dashboards with custom DAX measures that gave stakeholders real-time KPI visibility across multiple departments.
  • Handled the full ETL lifecycle — ingestion, transformation, and data-quality cleansing — across heterogeneous source systems.
AzureMicrosoft FabricPower BIDAXETL

Oracle APEX Developer Intern

Oradist · Tunis

Jun – Jul 2024
  • Built and shipped two internal business applications using Oracle APEX, replacing legacy manual processes used daily across the company.
  • Wrote PL/SQL stored procedures and triggers to enforce business logic and data integrity at the database layer.
  • Designed clean, role-based UI components that reduced onboarding and training time for end users.
Oracle APEXPL/SQLSQLUI Design

Billing & Logistics Agent

Centre National Pédagogique · Tunis

Summer 2023
  • Managed invoicing and billing operations — prepared client invoices, verified orders against stock, and maintained accurate financial records.
  • Prepared and packaged outgoing orders; handled full stock lifecycle including reception, storage, inventory counts, and reconciliation.
  • Counted and balanced daily revenue, flagging and resolving discrepancies before end-of-day closeout.
  • Served clients directly — assessed their needs and ensured timely fulfilment of educational material requests.
BillingInventoryClient RelationsLogisticsStock Management
Projects

Things I've built

From AI platforms to data pipelines and client e-commerce stores — end-to-end ownership on every project.

An end-to-end ITSM platform I architected and built from scratch as my 5-month final engineering project at Teamwill. The system brings genuine AI intelligence to IT support workflows — not just keyword matching, but deep contextual reasoning over ticket history. Key capabilities: - AI Chat Copilot: multi-intent chat with cause analysis, resolution steps, and evidence scoring. Handles problem listing, drill-down, and resolution guidance through a structured intent pipeline. - Auto-Classification: NLP-based ticket classification on draft creation; routes through a rules + ML engine to suggest category, priority, and assignee. - Proactive SLA Monitor: background service that fires alerts when tickets consume 75%+ of their SLA budget, preventing breaches before they happen. - Duplicate Detection: vector-similarity detection (threshold 0.72) to surface near-duplicate tickets before submission. - Recommendation Engine: Redis-cached AI recommendations with confidence scoring, feedback loops, and SLA resolution strategies. - Performance Analytics: agent performance dashboards, ticket insights, and classification logs for admins. - Redis Caching Layer: two-level cache (LRU + Redis) on embeddings and all major GET endpoints; 5-20 minute TTLs with smart cache busting on mutations.

Key highlights

  • LLM reasoning pipeline with multi-intent detection and evidence scoring
  • Two-level embedding cache: LRU + Redis SHA-256 key (24h TTL)
  • Proactive SLA alerting at 75% consumption threshold
  • Full Alembic migration chain across 36 schema versions
  • Built with AI coding agents (Claude Code) for architecture and velocity
FastAPINext.jsPostgreSQLRedisClaude AINLPSQLAlchemyAlembicDockerTailwind
Skills

Technical expertise

Built across real projects — not just tutorials. Each area reflects hands-on production use.

CC

AI-Assisted Engineering with Claude Code

I use Claude Code as a primary engineering tool — not a shortcut. It helps me navigate large codebases, reason about multi-service architecture, scaffold features precisely, and maintain quality across a system spanning FastAPI, Next.js, Redis, Alembic migrations, and a custom AI pipeline. This accelerates velocity while keeping the engineer in full control of every design decision.

AI & Machine Learning

LLM Integration (Claude / Anthropic API)90%
NLP & Intent Detection85%
Vector Embeddings & Semantic Search82%
Deep Learning (TensorFlow / scikit-learn)75%
n8n Workflow Automation78%

Backend & APIs

FastAPI + SQLAlchemy + Alembic92%
Spring Boot + Java78%
Node.js + Express.js80%
Redis Caching80%
PostgreSQL / SQL85%

Frontend

React + Next.js (App Router)88%
Tailwind CSS90%
TypeScript82%
MERN Stack80%

Data & BI

Power BI + DAX85%
Azure + Microsoft Fabric80%
SSIS + SQL Server75%
ETL Pipelines82%

DevOps & Tools

Docker78%
Git / GitHub88%
AI Coding Agents (Claude Code)92%
Jira + Figma75%

Creative & Media

Adobe After Effects80%
Adobe Premiere Pro82%
CapCut88%
Alight Motion78%

Tools & Technologies

PythonTypeScriptJavaSQLPL/SQLFastAPINext.jsReactSpring BootPostgreSQLMongoDBRedisOracle APEXAzureMicrosoft FabricDockerPower BISSISn8nClaude CodeTensorFlowscikit-learnAfter EffectsPremiere ProCapCutAlight Motion
Education & More

Background

Education

Bachelor's in Computer Science

Faculty of Sciences of Tunis (FST)

2023 – 2026

Scientific Baccalaureate

Honours

Lycée Sadiki, Tunis

2023

Certifications

NV

NVIDIA DLI

Fundamentals of Deep Learning

NVIDIA Deep Learning Institute

AW

AWS Academy

Cloud Fundamentals

Project-based

Activities

Free Software Club — FST

Member & Media Manager

2023 – 2025

  • Media Manager (2024–25): video editing, graphic design, live event coverage.
  • Co-organised two editions of the GENESIS Competitive Programming Contest.