Amazon Web Services
Amazon Web Services (AWS) offers a wide range of AI services, ranging from pre-built, out-of-the-box services that can make starting an AI project easier and minimize the need for experienced data scientists and AI developers. These services include, for example:
-
Amazon Translate (real-time translations),
-
Amazon Rekognition (image and video analysis),
-
Amazon Polly (Text-to-Speech) und
-
Amazon Transcribe (Speech-to-Text).
Managed infrastructure tools include:
-
Amazon SageMaker (create, train and deploy machine learning models),
-
Amazon Machine Learning drag-and-drop tools and templates (deploy ML models more easily),
-
Amazon Comprehend (Natural Language Processing),
-
Amazon Forecast (accurate time series forecasts) and
-
Amazon Personalize (personalized product and content suggestions).
In the area of generative AI, AWS offers:
-
Amazon Lex (creating AI chatbots),
-
Amazon CodeGuru (analyze and optimize code) and
-
Amazon Kendra (smart searches).
Microsoft
Microsoft’s Azure AI Services are aimed at developers and data scientists and are based on applications such as SQL Server, Office and Dynamics. The Redmond-based company has integrated artificial intelligence into various business apps – both in the cloud and on-premises.
As is known, Microsoft has entered into a close partnership with ChatGPT developer OpenAI. Accordingly, many AI applications can be found on the Azure marketplace. In addition, an OpenAI service with pre-trained large language models such as GPT-3.5, Codex and DALL-E 2 is available.
The ready-made AI services offered by Microsoft include:
-
speech recognition,
-
text analysis,
-
Translation,
-
Image processing and
-
ML-Modell-Deployment.
Google Cloud
Google’s AI service is focused on data analytics and offers tools such as BigQuery and AI Platform, as well as the AutoML service, which allows users with limited coding skills to automatically build models. Google Cloud also offers the Vertex AI platform to streamline AI workflows and simplify development and deployment. This also includes a wide range of services with pre-built solutions, custom model training and generative AI tools. With the Vertex AI Workbench, Google also provides a collaborative AI project environment for data scientists and developers.
Google Cloud’s pre-built AI solutions include:
-
Dialogflow (platform to develop chatbots and virtual assistants),
-
Natural Language API (Sentiment-Textanalysen, Entity-Extraktion etc.),
-
Vision AI (object recognition in images and videos),
-
Translation API (machine translations in various languages) as well
-
Speech-to-Text and Text-to-Speech (conversions between spoken language and text).
When it comes to Generative AI, Vertex AI Search and Conversation offers a collection of tools specifically designed to develop GenAI applications such as search engines and chatbots. This suite includes more than 130 pre-trained Foundational LLMs such as PaLM and Imagen to generate text and/or images. Google also has the AI assistant Gemini in its program, which is available in different versions.
IBM
IBM’s Watsonx, is a comprehensive AI tools and services offering known for its focus on automating complex business processes and its industry-specific solutions, particularly for healthcare and finance. Watsonx.ai Studio is the heart of this platform where you can train, validate, tune and deploy AI models – for both machine learning and generative AI. A data lakehouse ensures a secure and scalable storage system for your data (both structured and unstructured).
IBM’s AI Toolkit is a collection of pre-built tools and connectors that make it easier to integrate AI into your existing workflows. This allows you to automate tasks, gain insights from data and create intelligent applications. Watsonx also includes a number of pre-trained AI models that you can use directly without training. These models cover different tasks – for example:
Oracle
Oracle has lagged far behind the cloud hyperscalers so far, but it does bring some advantages that you should be aware of – primarily because it is a giant in business applications and databases. All applications installed on-premises can be moved to the cloud to achieve a hybrid configuration. This makes it much easier to move your local data to the cloud for data preparation and training purposes. Oracle has invested heavily in GPU technology, which is currently the primary means of AI data processing. So if you want to run AI applications on Nvidia technology, you can turn to Oracle. Another important advantage: Oracle’s AI services are among the cheapest.
Oracle Cloud Infrastructure (OCI) AI Services cover a broad portfolio of tools and services to provide companies with various AI functions. Similar to the case of IBM’s Watsonx, it is not a single service, but a collection of features that meet different needs – including:
Oracle’s Generative AI Services supports LLMs such as Cohere and Llama 2, enabling use cases such as:
-
writing assistants,
-
text summaries,
-
Chatbots or
-
Code generation.
Oracle’s Machine Learning Services provide tools for data scientists to build, train, and manage ML models. Popular open source frameworks such as TensorFlow and PyTorch are supported. Finally, with OCI Data Science, it is possible to deploy virtual machines with preconfigured environments for data science tasks – including Jupyter notebooks and access to popular libraries that simplify data exploration and model development workflows. (fm)
This article originally appeared at our sister publication Computerworld.com.
