Helicon - Radicalbit

One platform built for
Continuous Intelligence


What is Helicon


A simple, scalable, robust, code-free and generic platform to enable and productise the next generation of “online modified”, real-time event stream ML/AI models.

Simple, Secure and Scalable

Let our platform handle the underlying complexity. Rule your data workloads with novel awareness and watch it run at warp speed.

Unify training and evaluation

Continuous Training and Online Machine Learning features enable streaming inference and training, at once.

Human in the loop

The Organization is continuously involved in the process of models improvement through human or machine feedbacks.

Models improve when data change

Models are able to detect data and concept drift and adapt automatically decreasing the maintenance cost and increasing speed.

A suite of simple APIs to read and write data streams at scale, creating a powerful real-time experience

Bi-directional APIs that are flexible, scalable and easy to use. Helicon supports multiple development languages including Java, Javascript, Python. We also support also open protocols like MQTT, HTTP, GRPC so you can use the right technology based on your use case.

Automation via GUI

  • Read and Write data streams, at scale

    Bi-directional APIs that are flexible, scalable and easy to use

  • Integrate AI in seconds

    Deploy and Monitor your models effortlessly

  • Build pipelines with ease

    Design data processes and enrichment flows with a visual interface

Build and Manage streaming pipelines in hours instead of weeks

Simplify data engineering with the Helicon editor, no code needed to build complex pipelines to obtain fresh, high-quality real-time data.

Our editor avoids development and management pains and provides declarative pipeline development, testing and debug in one single place. This will help you focus on data correctness before deploying in production.

Govern your fast-moving workloads

Teams can easily implement business logic without spending time and effort in managing the infrastructure. The autoscaling data-aware resource can reduce the pipeline latency and maximize the resource performance. On pipeline performance, we can reduce the effect of hotkeys by automatically partitioning and constantly rebalancing data inputs.

CI/CD automatized with our APIs

AI projects are quite different from the DevOps tools and processes. Nevertheless, an approach based on continuous integration and deployment can be a benefit for AI projects.
With our full API, you can perform operations from many external management systems.

Top accurate models, forever

After an AI project has been successfully implemented, Helicon will start its magic. With its set of tools it will enable continuous training to detect drifts in data. Moreover, it will alert you when models are outdated or need to be re-trained.
Helicon it's based on NSDb, our time-series Database conceived having streaming real-time analytics in mind.

Next-gen real-time AI

Let your models learn from real-time data that changes over time and at different rates of speed. Helicon implements a completely new paradigm for training models while working online in the runtime environment, based on fresh data.