Data workflow development for object detection – Lyon, France
Headquartered in Lyon, France, Ryax Technologies is an early stage startup providing software that enables companies to industrialize their data science. The process of data science industrialization needs strong data engineering foundations. Complex tasks such as data analytics pipelines automations, hybrid infrastructure management, workflow scheduling, distributed systems configuration and operation, virtualized and containerized environment deployment, batch and stream processing workload orchestration, infrastructure and application monitoring along with optimizations of Big Data and AI frameworks are some of the data engineers’ responsibilities.
Our software platform, Ryax, implements the necessary data engineering plumbing by abstracting the underlying infrastructure and systems complexity to provide a platform with a simple to use interface for data scientists. It enables them to deploy their data analytics pipelines by focusing only on their expertise which is how to retrieve more business value from their data.
Object detection is of significant practical importance and can be used to answer a variety of questions. For example is an object present, how many objects, what is the type of the object, what is its size, where it is in respect to time. Therefore it can be used across a variety of industries such as manufacturing, augmented reality, autonomous vehicles, drone imagery, workplace automation and smart agriculture.
To perform automated object detection scientists often use the technique of deep learning which makes use of distributed computing and has proven particularly effective in hard and large-scale problems such as speech recognition, natural language processing, and image classification.
This internship will be focused on developing a typical data workflow for object detection along with the necessary integration within Ryax software to facilitate the usage of deep learning tools for object detection. The intern will develop within the Ryax software making use of its container (Docker) based environment bundling and deployment. At least one of these different known deep learning tools (Google Cloud AutoML, H2O, Spark deep learning pipelines, Tensorflow) will be integrated in Ryax and used within the workflow.
A realistic testbed will be configured using Raspberry Pi and Intel NUC gateways along with public or private cloud infrastructures.
The intern will work on the state of the art of object detection deep learning pipelines, she or he will develop in Python, R or Go and will make use of known open source tools such as Kubernetes, Docker, Tensorflow, Grafana, etc. After the developments, the intern will perform experiments on the designed testbed to validate the implementations, provide a performance analysis and describe possible paths for optimizations.
The intern should ideally have a data science background and must be confident in at least one language such as Python, R or Go. No previous usage of Kubernetes, SCADA or other tools are needed. However, experience with C, C++ or Java along with Docker containers and deep learning frameworks will be a plus.
Contact: Yiannis Georgiou: firstname.lastname@example.org
Duration – Compensation: 6 months – 577 Euros/month