Artificial intelligence(AI) and machine learning(ML) are growing in popularity in the technology ecosystem. Just a few months ago, it was relatively simple to keep up with the developments in the AI and ML markets. Today, that seems like a daunting task for most technologists as the space have been evolving incredibly fast.
The explosion of AI and ML platforms have created a very crowded market in which is very hard to distinguish signal from noise. However, despite the large number of AI technologies and startups, we can identify a few main categories that provide a good taxonomy to better understand the AI-ML markets. Let’s take a look:
Cognitive API Platforms
Cognitive API platforms abstract capabilities such as image, text and vision analysis using pre-trained AI models exposed as APIs.
— Some Platforms in the Space: Watson Developer Cloud, Microsoft Cognitive Services, Google Cloud, NLP, Speech APIs
— Good For: Performing well-defined tasks in areas such as text analytics, image processing and audio analytics.
— Not Great For: Implementing custom AI algorithms, training models…
Cloud ML Platforms
Cloud ML platforms are cloud services that abstract the creation and execution of ML applications.
— Some Platforms in the Space: Azure ML, Google Cloud ML, AWS ML.
— Good For: Creating cloud-first, large scale ML applications
— Not Great For: Integration with on-premise data sources, custom algorithms…
On-Premise ML Platforms
On-premise ML platforms enable the implementation of ML applications both on cloud and on-premise infrastructures.
— Some Platforms in the Space: Spark ML, Flink ML, H2O.ai
— Good For: Implementing custom algorithms, extensible and customizable ML applications.
— Not Great For: Simple ML scenarios, multi-language ML solutions…
Self-Service Data Science
Self service data science platforms provide interactive mechanisms for AI-ML applications.
— Some Platforms in the Space: Apache Jupyter, Apache Zeppelin, Google Cloud Datalab
— Good For: Integrative ML application, data exploration-visualization, human-centric ML solutions…
Deep Learning Frameworks
Deep learning frameworks are developer frameworks that provide the fundamental constructs to build deep learning applications.
— Some Platforms in the Space: TensorFlow, Torch, Caffe.
— Creating custom algorithm, customizable deep learning applications…
— Not Great For: Basic deep learning models [ex: image recognition, etc]…
Self-service ML Platforms
Self-service M Platforms allow non-developers to implement simple ML applications.
— Some Platforms in the Space: BigML, GE Wise.io
— Good For: Customizing well-defined ML models.
— Not Great For: Implementing complex ML solutions…
Vertical AI Solutions
Vertical AI solutions focus on the implementation of industry-specific, complex AI solutions powered by some of the aforementioned technologies.
— Some Platforms in the Space: IBM Watson, DeepMind.
— Good For: Implementing sophisticated, domain-specific AI solutions.
— Not Great For: Mainstream AI-ML models…
As a group, AI-ML marketplace technologies enable the cataloguing, management and monetization of AI-ML assets such as algorithms, data sets, etc.
— Some Platforms in the Space: Algorithmia
— Good For: Discovering and using custom AI algorithms
— Not Great For: Implementing end-to-end AI-ML solutions.