Oussama Makhbouche

Oussama Makhbouche is a Software Engineer and System Architect specializing in Applied AI, Clean Architecture, and Full-Stack Engineering. He currently works as a Junior AI & Data Analyst at POWR, where he architected and built the company's internal AI & Creator Intelligence Platform — featuring an enterprise RAG engine with pgvector, a campaign intelligence engine, an autonomous YouTube ad detection agent, and automated multi-platform ingestion pipelines. He is pursuing a Master's degree in Software Engineering at the Université de Montpellier (UM) and holds a Bachelor's degree in Software Engineering from King Saud University (KSU), Riyadh, Saudi Arabia, where he graduated as a merit-based scholarship recipient with a GPA of 4.85 / 5.00. His portfolio spans from low-level hardware integration (.NET/RFID) to high-level AI research (LSTM/RL) and modern web/mobile scaling — across 26+ public repositories.

Background

Oussama Makhbouche was born in Algeria. He completed his Baccalaureate with distinction, finishing 1st in Skikda and placing in the Top 15 nationally in Algeria — results that secured his merit scholarship for undergraduate study at King Saud University.

His interest in computing began early, driven by a curiosity for how large-scale systems are designed and how software can be engineered to be both reliable and impactful at a national level.

Education

Master's in Software Engineering — Université de Montpellier (UM)

Montpellier, France

• Program: M.Sc. in Software Engineering (Génie Logiciel).
• Focus areas: Advanced software architecture, distributed systems design, data engineering, and applied machine learning systems.

Bachelor's in Software Engineering — King Saud University (KSU)

Riyadh, Saudi Arabia · Merit-Based Scholarship Recipient

• GPA: 4.85 / 5.00
• Focus areas: Applied AI, Clean Architecture, distributed systems, hardware integration, and modern full-stack engineering.
• Notable coursework: Data Structures & Algorithms, Database Systems, Software Architecture, Machine Learning, Operating Systems, and Computer Networks.

During his studies, he has applied classroom theory to real-world engineering challenges — from architecting CLIMENCE, a national-scale air quality platform for the Ministry of Environment, to building GymAppManager, an enterprise desktop system with RFID/Turnstile hardware integration.

Professional Experience

Junior AI & Data Analyst — POWR

POWR · Riyadh, Saudi Arabia · 2026 – Present

Oussama designed, architected, and built internal AI and data infrastructure for marketing operations, turning fragmented creator and campaign data into a unified, queryable, AI-powered platform.

Creator Intelligence Platform (Internal AI Platform)

He designed and developed an internal AI-powered Creator Intelligence Platform that centralizes creator analytics, campaign intelligence, autonomous agents, and reporting for high-velocity marketing operations.

• Enterprise RAG & Semantic Search: Architected a production Retrieval-Augmented Generation (RAG) system with pgvector and OpenAI embeddings, combining vector search and SQL query synthesis to deliver instant natural-language answers over campaign databases, contracts, and creator history.
• Autonomous YouTube Ad Detection Agent: Built an AI agent pipeline that analyzes YouTube video content and transcripts to automatically detect, timestamp, categorize, and verify sponsored brand integrations and ad segments.
• Campaign Intelligence & Prediction Engine: Developed AI-powered modules for creator recommendation, campaign performance forecasting, audience overlap analysis, and automated budget allocation recommendations.
• Modular Backend Architecture: Architected a scalable FastAPI backend with PostgreSQL, SQLAlchemy, and Alembic, maintaining clean boundaries between ingestion, modeling, and presentation layers.
• Data Warehousing & Ingestion: Designed a three-tier data warehouse separating raw, staging, and warehouse layers to support scalable cross-platform analytics across Snapchat, YouTube, Kick, and Odoo.
• Production Dashboard: Built a high-performance React dashboard with TypeScript, TanStack Query, and Zustand for creator scouting, campaign tracking, real-time alerts, and performance reporting.
• Security & Governance: Implemented strict role-based access control (RBAC), thread management, audit logging, and automated test coverage across services using Pytest and Playwright.

Data Pipeline & API Integration

• API & OAuth automation: built an automated OAuth 2.0 authentication and token-refresh pipeline in Python/FastAPI, eliminating manual credential handshakes and securing long-term API access.
• Data pipeline engineering: engineered extraction scripts pulling live organic creator metrics — reach, story views, engagement, CTR, and snap detail — via the Snapchat Public Profile & Marketing APIs across multi-account structures.
• ETL & data transformation: developed transformation logic with pandas and openpyxl to map raw API JSON payloads into structured master tracking sheets and internal endpoints for real-time campaign auditing.
• System integration: configured FastAPI endpoints, local webhooks, and tunnel infrastructure to bridge live API metrics directly into database sinks and business tracking platforms.

Impact

• Consolidated siloed multi-platform creator metrics into a single source of truth powered by pgvector and RAG.
• Automated campaign audit reporting and ad verification, replacing manual tracking with continuous autonomous pipelines.
• Reduced influencer evaluation time from hours of manual vetting to instantaneous AI-assisted queries.

Stack: FastAPI · Python · PostgreSQL · pgvector · SQLAlchemy · Alembic · React · TypeScript · TanStack Query · Zustand · Playwright · Pytest · OpenAI APIs · RAG · AI Agents · Vector Search

Engineering Work

CLIMENCE — National-Scale Air Quality Platform (GovTech)

The flagship project of his portfolio, CLIMENCE is a cloud-native environmental intelligence platform integrating IoT drone telemetry and ESA Sentinel-5P satellite data through fault-tolerant ingestion pipelines. An LSTM-based forecasting engine generates multi-horizon AQI predictions for PM2.5, NO₂, and CO₂ across the Kingdom.

The system was designed and deployed on AWS with Docker containerization, FastAPI REST services, and a React visualization dashboard consumed by Ministry officials.

GymAppManager — Enterprise Gym Management (FinTech / System Design)

GymAppManager is an enterprise-grade desktop application built with .NET 10, C# 13, and EF Core 9 using strict Clean Architecture across five layers — Presentation, Application, Infrastructure, Domain, and Hardware. The system integrates RFID badge readers and turnstile hardware for physical access control, implements BCrypt password security, and features an offline-first automated backup system with configurable retention policies. This project demonstrates mastery of SOLID principles, Dependency Inversion, and zero-framework-leakage domain design.

VibeQuest — Educational RPG (EdTech / Gamification)

VibeQuest is an educational RPG simulation that gamifies learning AI-assisted development. Built with React 18, TypeScript, Zustand, and Framer Motion, it features dynamic curriculum scaling and a cyberpunk IDE aesthetic. Live →

JAR — Saudi Community Platform (Community / GovTech)

JAR is a Saudi neighborhood community app built with a product-minded engineering approach, focused on emergency response, marketplace exchange, events, and social interaction. Built with Flutter (Dart), Firebase (Auth/Firestore/Storage), and Google Maps API, it ships across iOS, Android, Web, and Desktop with Arabic-first RTL UX and a dedicated admin web panel. Live →

NAFAS — VR Exposure Therapy

NAFAS is currently an informational product website built with Vite, TypeScript, React, shadcn-ui, and Tailwind CSS, presenting the concept and clinical direction of the project. The full solution roadmap combines this web layer with Unity-powered virtual therapy rooms. It earned 2nd Place at a national-level hackathon. Live →

La Rose de Guelma — Commercial Website

La Rose de Guelma is a full-stack caftan e-commerce platform built with PHP 8.2, MySQL, Tailwind CSS, and Vanilla JavaScript in a monolithic architecture with a custom admin panel. It bridges inventory/database rigor with real Algerian logistics through a Wilaya/Commune delivery-pricing engine, advanced high-quality media handling, and 150+ iterative Railway deployments to refine performance and luxury UI responsiveness. Live →

Bridge — AI-Driven Educational Matching

Bridge uses a hybrid recommendation engine (collaborative filtering + content-based scoring) to match students with tutors, with delivery across both web and a Flutter/Dart mobile app. The system reduced time-to-match from manual search (~25 min) to under 10 seconds via algorithmic ranking validated through A/B testing.

Technical Skills

Languages — Python, C#, Java, TypeScript, JavaScript, Dart, SQL, C++, Go, Bash

AI & Machine Learning — TensorFlow, PyTorch, scikit-learn, LSTM and time-series forecasting, deep reinforcement learning, Retrieval-Augmented Generation (RAG), semantic and vector search, embeddings, AI agents, OpenAI APIs

Data Engineering & Analytics — pandas, NumPy, openpyxl, ETL pipeline design, layered data warehousing (raw / staging / warehouse), REST and OAuth 2.0 API ingestion, webhook-driven synchronization, data validation and auditing, Excel/CSV automation

Backend — FastAPI, .NET, Spring Boot, SQLAlchemy, Alembic, EF Core, PostgreSQL, MySQL, SQLite, Firebase, REST API design, role-based access control, audit logging

Frontend & Mobile — React, Vue, Nuxt, Next.js, TanStack Query, Zustand, Tailwind CSS, Flutter

Testing & Tooling — Pytest, Playwright, Git, Docker, AWS, Vercel, Figma

Architecture — Clean Architecture, SOLID, domain-driven design, modular service design, ingestion and event pipelines, hardware abstraction layers

Personal Life and Interests

Oussama is quadralingual: Arabic (native), English (C1 — PTE certified), French (C1), and Spanish (A2). He competes in swimming at the intercollegiate level, winning 3rd place in Saudi universities' swimming competitions.

His broader interests and current focus areas are shown in the table below: