A single platform to build deep learning models on
Create, evaluate, tweak and deploy deep learning models.
TRY FOR FREE
A single platform to build deep learning models on
Create, evaluate, tweak and deploy deep learning models.
TRY FOR FREE
  • Projects view of the Peltarion Platform
    Projects

    Projects view allows you to create a new project, select, edit or delete an existing project.

  • Datasets view of the Peltarion Platform
    Datasets

    Datasets view allows you to import, preprocess, augment and manipulate data to run experiments. The view gives a full overview of all the datasets available to the user, and gives access to information about the available datasets.

  • Modeling view of the Peltarion Platform
    Modeling

    The modeling view area allows you to build an AI deep learning model in one place. You can start from scratch or select from a series of preloaded snippets, build a model yourself or work with a colleague on the same one. There is no coding required, just drag and drop to build your model.

  • Evaluation
    Evaluation

    In the evaluation view, you can see in real time how the AI model is performing as it’s learning from the data. You can see the results of multiple experiments at the same time, so it’s easy to compare different models. The view lets you analyze in real time if the experiment is going in the right direction or not.

  • Deployment view of the Peltarion Platform
    Deployment

    When a model has been built and trained, it is time to deploy. Currently, you can deploy as API in the deployment view.

  • Projects

    Projects view allows you to create a new project, select, edit or delete an existing project.

  • Datasets

    Datasets view allows you to import, preprocess, augment and manipulate data to run experiments. The view gives a full overview of all the datasets available to the user, and gives access to information about the available datasets.

  • Modeling

    The modeling view area allows you to build an AI deep learning model in one place. You can start from scratch or select from a series of preloaded snippets, build a model yourself or work with a colleague on the same one. There is no coding required, just drag and drop to build your model.

  • Evaluation

    In the evaluation view, you can see in real time how the AI model is performing as it’s learning from the data. You can see the results of multiple experiments at the same time, so it’s easy to compare different models. The view lets you analyze in real time if the experiment is going in the right direction or not.

  • Deployment

    When a model has been built and trained, it is time to deploy. Currently, you can deploy as API in the deployment view.

Projects view allows you to create a new project, select, edit or delete an existing project.

Datasets view allows you to import, preprocess, augment and manipulate data to run experiments. The view gives a full overview of all the datasets available to the user, and gives access to information about the available datasets.

The modeling view area allows you to build an AI deep learning model in one place. You can start from scratch or select from a series of preloaded snippets, build a model yourself or work with a colleague on the same one. There is no coding required, just drag and drop to build your model.

In the evaluation view, you can see in real time how the AI model is performing as it’s learning from the data. You can see the results of multiple experiments at the same time, so it’s easy to compare different models. The view lets you analyze in real time if the experiment is going in the right direction or not.

When a model has been built and trained, it is time to deploy. Currently, you can deploy as API in the deployment view.

The platform allows me to explore and experiment quickly, solving problems and getting faster to solutions.

Daniel Skantze, Head of Engineering & Developer

Explore & experiment

  • No-code environment

    Build your AI models with standardized & modular drag-n-drop building blocks.

  • Pretrained networks

    Prebuilt & pretrained networks ready to work from. Run one or many in parallel.

  • Latest tech

    Use the latest deep learning techniques without needing to implement them yourself.

  • One-click deployment

    Deploy your models directly from the platform in one step.

Running an AI infrastructure with many different models in production, and being able to scale when I need. Never been easier.

Anders, PhD Artificial Intelligence & Head of Research

Infrastructure

  • No setup required

    Storage, computation and deployment resources are sorted and managed.

  • Train on powerful GPUs

    Get the resources that you need to train your applications.

  • Deploy on CPUs

    Run inferences on CPU-backed machines instantly.

  • Auto-scaling

    Scale models and deployments without any extra effort.

Carbonara? I’ve seen too much spaghetti code in my days. I like when things are in order. Ready. To. Go.

Jeff Eklund, Reliability Engineer

DevOps for deep learning

  • Version control

    Fully version controlled datasets, experiments, parameters, results and deployments, enable reproducibility, reusability, collaboration and governance.

  • Project history

    See the full history of your datasets, experiments and deployments, including when, who and where.

  • Full reproducibility

    Data, experiments, parameters, results and deployments are all under the same environment, so nothing is lost and results can always be reproduced.

  • Easy experimentation

    Easily build, configure, run and compare multiple experiments.

We see the real value when the AI models are in production. Being able to deploy, scale and run things quickly. Makes all the difference.

Liliana Lindberg, AI Solutions Engineer

Model deployment

  • Rest-API deployment

    Integrate model training and model predictions into your own applications and automate interactions with the platform with a few lines of code.

  • One-click deployment

    Deploy or rollback your models into production directly from the platform with one click. No extra work needed.

  • Deploy on CPUs

    Run inferences on CPU-backed machines instantly.

  • Scalability and availability

    Run any volume of predictions, thanks to autoscaled resources allocated to your applicationl.

Our clients’ goal is to be AI-first and distribute AI enablement throughout their entire organization. Then you really need to be able to collaborate smoothly.

Peder Ribbing, Head of Professional Services

Collaboration

  • Easy knowledge sharing

    Easily reuse previous work and pick up projects where others left off.

  • Better teamwork

    Easily prioritize ideas, reduce duplicated efforts and encourage collaboration.

  • Better insights

    Keep track of your team’s work volume, project progress and experiment metrics.

  • Keeping history

    Version control to enable reproducibility, reusability, collaboration and governance.

Latest product updates