Hi, I'm Amine Ouadaf

I'm a

Passionate about building intelligent systems and turning data into impactful solutions. I design and ship innovative, real-world software and AI-driven systems.

View my work Get in touch

6 Internships

30+ Technologies

3 Languages spoken

About Me

Learn, build, repeat.

Open to Software · Data · AI opportunities

Who am I?

An engineering student in Computer Science & Data Science at FSTM (Mohammedia), focused on Data & AI Engineering. I love building intelligent systems end to end — from data pipelines and ML models to the AI assistants and interfaces on top.

My Journey

I hold a Bachelor's in Mathematical Sciences & Applications from Cadi Ayyad University, which gave me strong foundations in mathematics and algorithms. Across six internships, I've applied them to data engineering, MLOps, AI and full-stack projects.

My Philosophy

I continuously sharpen my data and AI engineering skills to deliver reliable, real-world systems that turn data into measurable value.

Download my CV

Education

My academic background.

2024 — present

Engineering Degree in Computer Science and Data Science

Faculté des Sciences et Techniques, Mohammedia (FSTM)

Specialized training in computer science, data science, AI and software engineering.

2021 — 2024

Bachelor's Degree in Mathematical Sciences and Applications

Université Cadi Ayyad — Faculté Polydisciplinaire de Safi

Fundamental training in applied mathematics, analysis and algebra.

2020 — 2021

High School Diploma in Physical Sciences

Lycée qualifiant Al-Hidaya Islamia, Safi

Scientific training specialized in physics and mathematics.

Experience

Six internships across data engineering, MLOps, AI and full-stack development.

Jul — Aug 2026

Data & AI Engineering Intern — OCP Group

Safi, Morocco · On-site

Designed and built an end-to-end real-time data platform and multi-agent AI system for the supervision and predictive maintenance of an acid cooler in a sulfuric acid production unit.

Built a streaming ingestion chain (Apache Kafka → TimescaleDB) with Pydantic contract validation, physical-range checks, quality coding (good/suspect/bad) and a dead-letter table; designed a medallion architecture (Bronze/Silver/Gold) over TimescaleDB and a MinIO/Parquet data lake, orchestrated with Dagster (Polars batch jobs, Spark Structured Streaming aggregates).
Deployed online anomaly detection with River (Half-Space Trees) scoring each sensor stream in real time; benchmarked an offline Isolation Forest baseline in MLflow and versioned the online model through a champion alias — rollback without rebuild or redeployment; implemented concept-drift monitoring (ADWIN) with a Dagster sensor auto-triggering retraining.
Built a LangGraph supervisor exposed through FastAPI, orchestrating three specialized agents: an analyst (natural language → read-only SQL with allowlist validation), a documentalist (RAG over technical documentation and incident history with Qdrant and locally computed embeddings) and a diagnostician correlating anomalies, drift, data and docs.
Instrumented observability with Langfuse (traces, costs, evaluations) over a provider-agnostic LLM layer (Anthropic, Groq, Ollama) with a deterministic evaluation set replayed in CI; developed a React/TypeScript web app (real-time dashboards, alerting, conversational interface) packaged as a single Docker Compose stack deployed on a VPS.

Jul — Aug 2026

AI & Backend Engineering Intern — Voynich

Mohammedia, Morocco · Remote

Built an AI agent that monitors Moroccan public procurement portals and gives each company a ranked, explained shortlist of the tenders actually worth bidding on — replacing a daily manual check of several portals.

Built a per-portal scraping layer behind a shared BaseSource contract (httpx/BeautifulSoup) with rate limiting, deduplication and French date parsing — adding a portal touches a single file; orchestrated the pipeline as idempotent, retryable Prefect flows (scrape → understand → score → alert → notify) running unattended on a schedule.
Designed a provider-agnostic LLM client (OpenAI, Anthropic) for structured extraction from raw announcement text, and built the relevance scorer: every tender × company pair scored 0–100 with a one-sentence justification, with caching and profile fingerprinting so nothing is recomputed on unchanged data.
Designed on-demand ingestion of consultation files (DCE) for high-scoring tenders only, stored in S3-compatible object storage, with text extraction over the archives and a per-tender chatbot answering questions on their content; shipped dashboard alerts, deadline reminders and an email outbox with per-company preferences, daily digests and retries.
Co-built the multi-tenant FastAPI + MongoDB backend and the React/Tailwind dashboard, where editing a company profile instantly re-ranks its shortlist; containerised and deployed the whole stack on Railway (API, flow worker, Prefect server, its Postgres and the React UI).

Apr — Jun 2026

Data & BI Engineering Intern — InnuCloud

Montreal, Quebec, Canada · Internship

Designed and built an end-to-end open-source modern data platform based on a lakehouse architecture for an IT services company, covering the full data lifecycle from ingestion and real-time streaming to data modeling and BI reporting.

Developed ETL/ELT pipelines from heterogeneous sources (files, databases, REST APIs, streaming) and implemented real-time streaming with Apache Kafka and Schema Registry for ticket lifecycle events and infrastructure alerts.
Modeled data with dbt Core (staging, marts, SCD Type 2) with 180+ automated data quality tests; orchestrated workflows with Apache Airflow (DAGs, dependencies, SLA monitoring); ensured data quality with dbt-expectations, Great Expectations, Prometheus and Grafana.
Designed a data lake on MinIO (S3) with medallion architecture (Bronze/Silver/Gold) and Apache Iceberg lakehouse (time travel, versioning) queried via Trino; optimized analytics with ClickHouse; built Power BI dashboards connected to PostgreSQL and ClickHouse with DAX measures.

Sep 2025

Software Engineer Intern — SecurePath Consulting

France · Remote

Contributed to a cyber threat intelligence platform, focused on data integration and quality.

Integrated and enriched threat data from multiple trusted sources.
Designed an LLM-based pipeline to detect and manage duplicate threat data, improving data quality and efficiency.
Built unit/integration tests and Swagger API documentation, improving maintainability and developer adoption.

Aug 2025

AI Software Developer Intern — OCP Group

Safi, Morocco · On-site

Built an AI-powered IT helpdesk chatbot to automate internal support.

Designed the chatbot using RAG, NLP intent classification and OpenAI models.
Architected a microservices system with WebSockets, REST APIs and secure JWT/OAuth authentication.
Integrated Jira REST APIs to manage the support ticket lifecycle, reducing response time.

Apr — Jul 2025

Full-Stack Developer Intern — NextMindz

Casablanca, Morocco · Hybrid

Architected a scalable full-stack, AI-powered CRM for freelancers and SMEs.

Built the CRM with React, TypeScript and Supabase — modular architecture and responsive UI.
Implemented core SaaS features: auth, role-based access, Kanban sales pipeline, PDF invoicing and Stripe subscriptions.
Integrated a RAG-based assistant so users can ask questions about their data and get contextual answers.

Skills

My everyday toolkit.

Languages

AI & Machine Learning

Backend & APIs

Frontend

Databases

Data Engineering

DevOps & Cloud

Projects

A selection of things I've built.

OCP Group · Data Engineering, MLOps & Agentic AI

Real-Time Platform & Multi-Agent AI for Predictive Maintenance

End-to-end real-time data platform and multi-agent AI system supervising an acid cooler in a sulfuric acid production unit: streaming ingestion, medallion lakehouse, online anomaly detection with drift-triggered retraining, and a LangGraph agent team answering operators in natural language.

Voynich · Agentic AI & Backend

AI Agent for Public Procurement Monitoring

An AI agent that monitors Moroccan public procurement portals and gives each company a ranked, explained shortlist of the tenders actually worth bidding on — replacing a daily manual check of several portals.

Personal project · Azure Data Engineering

End-to-End Azure Retail Data Pipeline (Medallion)

1.07M retail transactions through a Bronze/Silver/Gold medallion architecture on Azure: metadata-driven Data Factory ingestion, PySpark on Databricks, Delta Lake with Unity Catalog, Synapse Serverless and Power BI, on Terraform-provisioned infrastructure.

View code

InnuCloud · Data Engineering & BI

Open-Source Modern Data Platform (Lakehouse)

Designed and built an end-to-end open-source modern data platform based on a lakehouse architecture for an IT services company, integrating multi-domain data (consultant activity, support tickets, infrastructure metrics, HR, and billing).

SecurePath Consulting · AI & Data

Cyber Threat Intelligence Platform

Contributed to a cyber threat intelligence system by integrating and enriching data from multiple trusted sources, with an LLM-based pipeline to detect and manage duplicate threat data.

OCP Group · AI & Microservices

AI-Powered IT Helpdesk Chatbot

Built an AI-powered helpdesk chatbot using RAG, NLP intent classification and OpenAI models for intelligent IT support automation, integrated with Jira for ticket lifecycle management.

NextMindz · Full-Stack & AI

AI-Powered CRM Platform

Architected a scalable full-stack AI-powered CRM with React, TypeScript and Supabase, featuring secure auth, role-based access, Kanban pipeline, PDF invoicing, Stripe subscriptions and a RAG-based assistant.

Microservices · Data Engineering

Smart Energy Management System (SEMS)

IoT energy platform with real-time collection, analytics and anomaly detection over millions of sensor readings. 7 Spring Boot microservices with Kafka messaging, API Gateway, Eureka discovery and resilience patterns.

Full-Stack ML Pipeline

E-commerce Spending Prediction

End-to-end ML pipeline predicting customer spending: a FastAPI service with prediction & batch endpoints and data validation, plus an interactive Streamlit app with confidence-interval analysis.

Get In Touch

A project, a question, an opportunity? Drop me a line.

Location

Morocco

Email

amine.ouadaf683@gmail.com

Phone

+212 6 59 72 66 08