About
Hi, I'm Adem.
Robotics and electronics by training, software and MLOps by trade. I've shipped web and mobile apps, real-time video analysis, autonomous-store systems and GPU model serving. Now I'm building my own products, and telling the whole story here.
- Based in
- Paris, France
- Currently
- Staff MLOps Engineer, Integral Ad Science
- Experience
- 10 years
01Experience
The journey so far
Staff MLOps Engineer
Integral Ad Science · Employee · Paris, France
Development and maintenance of Total Media Quality (TMQ), a real-time data processing pipeline for online advertising.
- Identified the GPU bottlenecks and isolated them into a dedicated service
- Designed a model-serving architecture to scale GPU resources, implemented with Nvidia Triton
- Aligned CI/CD with IAS deployment constraints and added functional and integration tests
- Improved observability with Prometheus and Grafana, and proposed alerting on performance and latency anomalies
PythonFastAPIPyTorchNvidia TritonKubernetesAWSKafkaDatabricksMLflowPrometheusGrafanaTech Lead / Senior Software Engineer
Freelance · Freelance · Hybrid, Paris
Software engineering expertise (backend, DevOps, frontend, machine learning) for several companies.
- Ondorse: development and maintenance of a KYB (Know Your Business) SaaS
- NielsenIQ: development of ScanRewards, a mobile receipt-collection app
- AnotherBrain: monitoring stack for the Phosphor Quality product
- Context: benchmark of Ray, a Python framework for distributed ML
PythonFastAPIPostgreSQLReactTypeScriptCeleryRayKubernetesPrometheusGrafanaSoftware Engineer
Boxy (formerly Storelift) · Employee · Ivry-sur-Seine, France
In the Store-AI team, design and improvement of the applications of Boxy, a fully autonomous connected store.
- Added real-time event tracking to monitor customer behaviour, with logging and reporting for business intelligence
- Integrated temperature sensors and built an IoT app for cloud temperature monitoring
- Designed the microservices architecture of Boxy v2
- Set up code quality procedures (linting, tests, static typing) automated in GitLab CI and Git hooks
PythonKubernetesMQTTRabbitMQRaspberry PiArduinoFluentdKibanaGitLab CISoftware Engineer
Context · Employee · Paris, France
Design and development of software around real-time video analysis with deep learning.
- Built an internal tool to manage image and video datasets, and modernised the annotation tools
- Designed a REST API for video upload and analysis (celebrity, logo and object detection), with versioning and blue-green deployments
- Developed a client dashboard for video analytics
- Refactored the core compute service and moved to microservices on Kafka, deployed on Kubernetes
PythonPyTorchDjangoFastAPIReactAngularPostgreSQLKafkaKubernetesAWS
02Education
Robotics, electronics and signal
- with honors
Master 2, Image & Sound for Intelligent Systems
Pierre and Marie Curie University (Paris VI)
- high honors
Master 1, Robotics Engineering & Intelligent Systems
Pierre and Marie Curie University (Paris VI)
- high honors
Bachelor, Electronic Engineering
Pierre and Marie Curie University (Paris VI)
- honors
Baccalaureate, Science & Engineering
Le Corbusier High School (Aubervilliers)
03Skills
What I work with
Software
Agentic AI
Hardware
Robotics
04Languages
Three languages
- French
- Native
- Turkish
- Native
- English
- Fluent
05Certifications
Keeping the theory sharp
- Certificate
Machine Learning Specialization
DeepLearning.AI, Stanford University
- Certificate
Deep Learning Specialization
DeepLearning.AI, Stanford University