ST

Backend · AI · Full-Stack

Saad Tachrimant

Curriculum vitae

Backend & ML Systems Engineer

Three-plus years in Java and Spring Boot, having shipped two SaaS platforms from domain modelling through production deployment — one an ERP now running across 10+ public university hospitals — alongside two years of applied ML: pipelines from ingestion to inference, distributed federated learning across edge hardware, and time-series forecasting on large-scale renewable energy data. Peer-reviewed publication at ACM e-Energy 2025.

The combination matters because a model only earns its keep with the service engineering, data plumbing, and deployment discipline around it.

Open to backend, ML and full-stack roles across Europe, Canada and the US — available for relocation or fully remote work.

Java 21 / Spring Boot Python / PyTorch Microservices / DDD Kafka PostgreSQL Federated learning Docker / Kubernetes
Download PDF CV Contact me

tachrimantsaad@gmail.com

+212 608 736 429

Marrakech, Morocco · open to relocation

saadtachrimant.com

Experience

Professional experience

Research Engineer

UM6P – College of Computing · Benguerir, Morocco · 11/2024 – Present

  • Designed and deployed end-to-end machine learning pipelines — data ingestion, training, and inference — running on production IoT systems, with domain logic kept separate from infrastructure so models could be swapped without rewriting the surrounding service.
  • Built a distributed peer-to-peer federated learning framework across a Raspberry Pi cluster with event-driven node coordination, fault tolerance under node dropout, and asynchronous FedAvg weight aggregation with α-weighted sharing between peers.
  • Developed LSTM and ARIMA forecasting models for large-scale renewable energy generation, and ran a full degradation sweep showing that fixed statistical models decay at roughly +0.85 pp RMSE per year as grid penetration rises, while an expanding-window LSTM stays bounded.
  • Engineered preprocessing pipelines for datasets carrying 38% missing values and 18% outliers, turning unusable raw sensor data into a reliable training corpus.
  • Debugged distributed training failures across nodes, including scheduler misconfiguration, silent hyperparameter fallback, and a feature-dimension mismatch that corrupted aggregation without raising an error.
  • Co-authored the THERMODSET dataset paper published at ACM e-Energy 2025, with reproducible experiments and deployment-oriented design.

Python · PyTorch · pandas · NumPy · scikit-learn · Java · Docker · PostgreSQL · REST APIs · Raspberry Pi · Federated Learning

Backend Engineer

EnovaR&T · Rabat, Morocco · 04/2023 – 11/2024

  • Designed and built a hospital management ERP SaaS deployed across 10+ Moroccan public university hospitals (CHUs) and integrated with 50+ partner clinics and pharmaceutical laboratories, covering hospitalisation, emergency, pharmacy, surgical blocks, and billing, with each domain implemented as an isolated microservice using DDD-based modelling.
  • Built a configurable workflow engine in hexagonal architecture, letting business rules be reconfigured per hospital client without touching core code; removed race conditions on bed availability with a reservation-then-confirm pattern.
  • Developed Spring Boot REST APIs with unit and integration test coverage, running in live clinical environments.
  • Implemented authentication and authorisation with Spring Security and JWT, enforcing role-based access control across clinical, nursing, and administrative user roles in every domain.
  • Owned PostgreSQL schema design and Flyway migrations; eliminated N+1 queries and optimised slow queries under production load.
  • Containerised and deployed services with Docker across environments through a Jenkins CI/CD pipeline.

Java · Spring Boot · Spring Security · PostgreSQL · Flyway · Docker · Jenkins · Angular · Microservices · DDD

Full Stack Engineer (Freelance)

Z Smart Services · Marrakech, Morocco · 08/2021 – 04/2023

  • Built and shipped a data labelling platform for image, text, and video datasets, used in production ML annotation workflows.
  • Designed and built a secure e-learning platform with a Spring Boot REST API, JWT authentication, and Angular frontend.
  • Shipped 3 client projects end to end, from requirements analysis through production deployment; all reached live use.

Java · Spring Boot · Python · Flask · MySQL · MongoDB · Angular · Vue.js

Publications

THERMODSET: A Thermal Dataset for Building Energy Modelling

ACM e-Energy · 2025

DOI: 10.1145/3679240.3734644

Education

MSc – Big Data & Cloud Computing

Ibn Tofail University · Kenitra · 12/2023

BSc – Distributed Information Systems

Cadi Ayyad University · Marrakech · 06/2021

Languages

Arabic — native

French — fluent (B2 certified, 2021)

English — fluent

Spanish — beginner

German — beginner

Skills

Machine Learning

PyTorch, scikit-learn, pandas, NumPy, time series forecasting, LSTM, ARIMA, federated learning, FedAvg, model training, inference, feature engineering, data preprocessing, experimental design.

ML Systems & Pipelines

ML pipeline design, data ingestion, training pipelines, edge inference, model deployment, reproducible experiments, distributed training.

Languages

Python, Java 21, Java 17, SQL, TypeScript, JavaScript.

Backend & APIs

Spring Boot, Spring Data, Spring Security, REST API, Flask, microservices, API design.

Data & Streaming

PostgreSQL, MySQL, MongoDB, Apache Kafka, RabbitMQ, Redis, event streaming, schema design, Flyway, query optimization.

Architecture

Distributed systems, event-driven architecture, Domain-Driven Design (DDD), hexagonal architecture, SOLID, fault tolerance.

Infrastructure

Docker, Kubernetes, Jenkins, GitLab CI, GitHub Actions, CI/CD, Linux, ARM / Raspberry Pi edge devices, Maven, Git.

Testing & practices

JUnit, Mockito, unit testing, integration testing, TDD, test coverage, Agile, Scrum, code review, technical writing.