Model Evaluation Jobs

I have a private collection of images that I want to sort automatically, and I need your expertise to build and fine-tune an image-classification pipeline. The dataset is purely visual—no accompanying text or numerical metadata—so the focus is on extracting meaningful visual features and assigning each image to the correct category. My end goal is straightforward: a reliable model that can classify every new photo I add, with clear metrics that show its accuracy. I haven’t fixed the exact number of categories yet; we can review the sample set together and decide what makes sense. You are free to work in Python with libraries such as TensorFlow, Keras or PyTorch, and any supportive tools for preprocessing, augmentation, and evaluation. I’ll share the image folder...

Je souhaite concrétiser un projet résolument axé Technologies, plus précisément sur l’Intelligence artificielle. L’idée est d’exploiter tout le potentiel de l’IA pour créer une solution utile, fiable et évolutive : de la définition du cas d’usage jusqu’au déploiement d’un prototype fonctionnel. Votre mission : – M’aider à préciser la problématique métier qui tirera vraiment parti de l’IA. – Concevoir l’architecture (data, modèles, pipeline MLOps) adaptée à cette problématique. – Développer puis tester un premier MVP, avec documentation claire pour une prise en main rapide. Je ...

I’m sitting on a very large, fully structured dataset and need an expert who can turn that raw volume into reliable, actionable insight. The core of the engagement is data analysis—everything from designing an efficient pipeline that ingests, cleans, and joins the tables at scale, to writing performant queries and statistical routines that surface trends I can use in decision-making. The stack is open: Apache Spark, Hadoop, distributed SQL engines, or another platform you’re confident will keep runtimes reasonable and costs under control. What matters most is that the solution handles billions of rows without sacrificing accuracy or maintainability and that every step you take is reproducible. Acceptance for the work will be based on: • A documented workflow (cod...