BlackFabric

Applying AI in your organisation can be challenging. BlackFabric helps you address this business challenge with a centralised AI/ML platform.

Accelerate your company’s digital transformation with a platform that allows your developers and data scientists to build, train and deploy machine learning models faster.

BlackFabric’s AI/ML engine has been built with the TensorFlow framework and scaling in mind, providing an end-to-end pipeline for training and inference tasks. Components are created in separate containers, allowing for deployment in container-orchestration tools such as Kubernetes, Docker Swarm, ECS, AWS Fargate or Azure Containers Services. 

BlackFabric is omni-cloud and hot pluggable into your applications, data sets or IoT devices.

How it works

Label

Prepare and store your data sets. Label your training data.

Build

Build your own models or leverage pre-built algorithms and models.

Train

Train and validate your models in multiple pipelines.

Deploy

Prepare and store your data sets. Label your training data.

Manage

Build your own models or leverage pre-built algorithms and models.

How it works

Label

Prepare and store your data sets. Label your training data.

Build

Build your own models or leverage pre-built algorithms and models.

Train

Train and validate your models in multiple pipelines.

Deploy

Prepare and store your data sets. Label your training data.

Manage

Build your own models or leverage pre-built algorithms and models.

Functional overview

Visualisation

Present data in your preferred BI tools (Power BI, Tableau, etc.)

Machine Learning models

Put ML into action with pre-built models that can be used within minutes or build your own algorithms and upload them directly into BlackFabric.

Computer vision

The object detection model allows you to identify objects in images and videos. Pipe AI is a sub-set of this component, allowing water utilities organisations to detect pipe defects and anomalies.

Prediction

The prediction component allows you to apply different time series forecasting models to any data set and generate forecasts/predictions. This is particularly useful for IoT data.

Anomaly

The anomaly component allows you to ingest both labelled (supervised learning) and unlabelled data sets (unsupervised learning) to look for any significant anomalies or clusters of anomalies that may be of interest to the use case.

Custom model

The custom model component allows you to utilise your own models, or any open source model architectures that are not currently included off the shelf with BlackFabric.

Computer vision

The object detection model allows you to identify objects in images and videos. Pipe AI is a sub-set of this component, allowing water utilities organisations to detect pipe defects and anomalies.

Prediction

The prediction component allows you to apply different time series forecasting models to any data set and generate forecasts/predictions. This is particularly useful for IoT data.

Anomaly

The anomaly component allows you to ingest both labelled (supervised learning) and unlabelled data sets (unsupervised learning) to look for any significant anomalies or clusters of anomalies that may be of interest to the use case.

Custom model

The custom model component allows you to utilise your own models, or any open source model architectures that are not currently included off the shelf with BlackFabric.

Pipelines

Our AI/ML engine offers several pipelines that can support multiple training and inferencing tasks simultaneously.

Training pipeline

Learn a new capability from existing data.

Inference pipeline

Apply the capability to new data.

Reinforcement learning pipeline

Run simulations.

Closed learning pipeline

Allow for closed loop systems.

Training pipeline

Learn a new capability from existing data.

Inference pipeline

Apply the capability to new data.

Reinforcement learning pipeline

Run simulations.

Closed learning pipeline

Allow for closed loop systems.

Data curation, cleansing & analysis

Turn independent data sources into unified data sets for analytics. Our process involves curation, augmentation, cleansing and analysis.

Data aggregation & brokering

Use of Databricks to ingest data from multiple sources and provide conditioned data sets.

Data storage

Designed to work omni-cloud or on-premise, you can use Oracle, SQL, AWS, Azure or Google Cloud to stage data for faster processing.

Become an AI driven organisation today

Plug in ML models into workflows

Deploy and integrate your models into workflows and systems so actions can be taken on model decisions.

Infuse AI in your business applications

Deploy and integrate your models into workflows and systems so actions can be taken on model decisions.

Launch your own AI-powered application

Deploy and integrate your models into workflows and systems so actions can be taken on model decisions.

Pipe AI

Pipe AI

Part of BlackFabric is Pipe AI, an asset management platform powered by machine learning and computer vision. As a game-changer for the engineering and construction industry, Pipe AI uses artificial intelligence to predict maintenance and failures, and review CCTV footage to identify cracks and anomalies in pipes.​