The latest edition of AWS Reinvent in 2023 generated a lot of interest among cloud providers thanks to major announcements around AI, and revolutionary tools such as Expert Assistant and Business Expert. We take a closer look at these innovations, which promise to transform the cloud landscape in 2024.
By putting AI at the heart of the transformation, AWS Reinvent 2023 confirms its commitment to catching up by offering a credible alternative to Microsoft solutions with a customer-centric approach. The combination of "generic" and "AWS" AI products offers a convincing solution, with the Amazon Q suite at the forefront.
The Amazon Q revolution: redefining the AWS experience
Amazon Q, the new family of services focused on generative AI, will provide users with specialised assistants for the various domains supported by AWS. Comprising four distinct products, it aims to facilitate adoption of the cloud.
- Amazon Q Expert Assistant : An indispensable ally for AWS architecture design, incident resolution and the integration of best practices. A major asset for simplifying the use of AWS products.
- Amazon Q Code Transformation : With a focus on development, it automates the migration of Java code when Java Development Kit (JDK) versions are updated, with the eventual aim of migrating Windows applications to Linux.
- Amazon Q Business Expert : It's an assistant that brings together data from several sources such as S3, Salesforce, MS365, Service Now, Google Suite, Atlassian and more. This makes it an ideal answer aggregator for companies with multiple technologies. A real plus for knowledge sharing within companies.
- Amazon Q Connect: It specialises in user support. It is the ideal solution to help call centres respond effectively to customer queries by generating automatic responses.
Beyond Q: other impactful announcements
AWS also revealed major announcements such as the new Amazon Bedrock model, which offers several specialisations depending on the use case. The models include Claude d'Antropic, LLama 2 (Open Source by Meta), and Stability AI, opening up Fine Tuning opportunities for additional training based on the data available.
AWS offers pre-trained LLMs with the added option of teaching them highly contextualised elements. An interesting opportunity for research and highly specialised professions that want to take advantage of AI to make it learn specific elements.
Security and confidentiality: a major challenge for AWS
Data security and confidentiality, particularly in relation to AI, remain a major issue for AWS. AWS has announced that the data used via Bedrock will remain private and secure. This is a crucial point at a time when data sovereignty at European and international level is a key concern.
New chips: Towards the democratisation of AI?
Amazon has unveiled a new generation of Graviton4 and Trainium2 chips, which promise faster, cheaper and less power-hungry AI training, making AI accessible to a wide range of users.
Key facts:
These announcements have the merit of reassuring AWS users about the firm's desire and current progress on AI. Amazon has clearly jumped on the AI bandwagon, and in a big way! We can't wait to get our hands on these new solutions and test their level of maturity against the technologies offered by Microsoft (ChatGPT, Copilot, etc.) and Google with its Gemini.