CLOUD ARCHITECTURE · BACKEND ENGINEERING · OBSERVABILITY · APPLIED AI
From signal to system. We design, build and operate the backends that keep critical platforms running.
Thina Technologies is an independent engineering studio led by a PhD in statistical signal processing who has spent close to two decades shipping production systems: event-driven microservices, multi-cloud infrastructure, and the observability that keeps them honest.
statistical signal processing & Bayesian inference
AWS · Azure
multi-cloud delivery, infrastructure as code
Trainer
Kafka, WildFly, Spring & JEE microservices
01 — EXPERTISE
Six disciplines, one goal: systems that scale and that you can see into.
Cloud architecture
Target architectures, migrations and landing zones on AWS and Azure — designed for cost, resilience and auditability, and delivered as code with Terraform.
REST and SOAP APIs, domain-driven microservices and secured integrations — from legacy JEE modernisation to cloud-native Spring Boot and Kotlin services.
Java · Kotlin · Spring Boot · Spring Security · JEE · OAuth2 · Keycloak · WildFly
Event streaming & data
Event-driven backbones and search platforms that carry high volumes reliably — topic design, stream processing, index lifecycle and storage strategy.
End-to-end tracing of HTTP exchanges and business flows, searchable audit trails, alerting and the operations work that keeps clusters healthy under load.
Retrieval-augmented search over technical and audit data, document-understanding pipelines and agent deployments — grounded in a statistics and machine-learning background.
LangChain4j · pgvector · Amazon Bedrock · Vertex AI · Google ADK · TensorFlow
Technical leadership & training
Architecture reviews, technical leadership inside client teams, and hands-on courses that leave your engineers autonomous.
Instrumentation for what your services actually do.
Kockpit is an open-source engineering platform for Spring Boot systems. It audits every call your services make in and out, lets you change features and configuration without redeploying, and gives you one web console to watch it all.
→Audit trail — HTTP, batch and key-value events, rule runs and OpenAPI operations, with sensitive fields masked in JSON and XML
→Feature flags & dynamic configuration — changed live from the console, with history
→Caches, instance health & manifests — per-app cache stats, heartbeats, domains and environments
→Rule engine — its own DSL, a Maven plugin and a visual designer
→AI-ready — an MCP server on Spring AI lets AI assistants search your audits
Trusted on platforms where downtime is not an option.
HOSPITALITY · 2018 → today
Accor
Architecture & backend for the Web Commerce Platform
REST and SOAP APIs and microservices behind Accor's booking and commerce ecosystem, deployed on AWS, plus audit and tracing of the platform's exchanges.
With the team, delivered a complete rework of the customer referential (RCU) on Azure, and now building a new AI-agent solution on Google Cloud with Google's Agent Platform and Gemini.
Azure · Google Cloud · Agent Platform · Gemini · AI agents
DIGITAL MAIL · 2007 → 2020s
Maileva · Docaposte
Architect & technical lead
Led the move from JEE monoliths to containerised microservices for La Poste group's digital-mail platform: API design, Kafka backbone, OAuth2 security and Kubernetes delivery.
WildFlyadministration, clustering & high availability
Spring & JEE microservicesdesign, security, containerisation
05 — ABOUT
Mahieddine M. Ichir, PhD
Founder & President, cloud architect and software engineer
I started in research — a PhD in statistical signal processing and Bayesian inference at Université Paris-Sud and the L2S–CNRS laboratory — and moved into software in 2007. Since then I have gone from C on Linux to JEE, from monoliths to microservices, and from data centres to multi-cloud.
That research training still shapes how I work: measure first, model the system, then build. It is why observability sits at the centre of everything Thina Technologies does, and why Kockpit exists.
EDUCATION
PhD, Signal & Image Processing — Paris-Sud XI / L2S CNRS, 2005
MSc, Automatic Control & Signal Processing — Paris-Sud / Supélec, 2002
Electrical & Electronics Engineer — INELEC, 2001
CERTIFICATIONS
Neural Networks and Deep Learning — Coursera
TensorFlow for AI, ML and Deep Learning — Coursera