Not all industry trade shows are born equal. Trying to discern which to attend can be a bit of a minefield – but AWS re:Invent is always one to watch. A key event in the tech industry calendar, it’s one of the biggest tech learning conferences globally, and this year’s event did not disappoint. It featured over 2,500 tech talks, was attended by a massive 65,000 tech professionals, oh and a crack team of QA-ers were there too. We gave talks, chatted to thousands of people on our stand, attended as many speaker slots as we could and took copious notes to report back. So if you missed it – don’t worry – we have you covered.
The 10 key things to take away:
1. AWS launched A LOT of new products this year
AWS launched 77 products, feature releases and services at re:Invent 2019 – with the most launches in machine learning (where 20 announcements were made). It’s clear AWS want to remain ahead of the curve, supporting customers’ with transformation and covering all bases. They are developing their products to make using the cloud easier, faster and more accessible. Among the announcements were ways to automate or speed up processes – and tackle some big customer issues to remove barriers to transformation. The market is increasingly looking to get more and more value from big data with machine learning, and are going to extraordinary lengths to achieve this – so AWS have focused the most energy and innovation on making machine learning easier. They’ve created the broadest and deepest set of machine learning and AI services to put machine learning in the hands of every developer.
2. Investing in training and cloud skills is crucial when transitioning to the cloud.
The event keynote unpacked barriers businesses face when transforming to the cloud and identified ways to address this. Equipping developers with a broad range of capabilities and investment in training was identified as key to driving this transformation . Another helpful piece of advice that was shared was not to let paralysis stop you before you start. Prioritise workloads to transform now, and what can transform longer term. Don’t be held back by thinking everything has to transform at once.
3. Barriers to transformation have been removed with AWS infrastructure on-premises.
A big barrier businesses have faced to cloud transformation is the requirement for some workloads to remain on-premises. At ReInvent AWS has launched Outposts – AWS infrastructure you can have on-premises which connects to all other AWS applications. This means that IT teams can reduce the time, resources, operational risk, and maintenance downtime required when managing IT infrastructure.
4. Serverless containers are in demand.
AWS announced that it has expanded its serverless compute engine for containers, Fargate, to run Kuberenetes applications. They announced that Fargate has become a broadly demanded service, and that 40% of new AWS customers start with it. Fargate makes it straightforward to run Kubernetes-based applications on AWS by removing the need for the provisioning and management of infrastructure for pods.
5. Having the right tool for the right job is imperative. Achieve it with purpose-built databases.
Organisations are moving away from relational databases to purpose-built databases. AWS have continued expanding their range of purpose-built databases to include a Cassandra-compatible managed service. AWS now has a total of 14 different purpose-built databases addressing many customer needs.
6. Optimise performance to work faster with big data.
Many of the updates were around improving performance and latency to analyse big data. There are updates to Redshift to create faster performance and support complex queries. They’re also launching AQUA, a query accelerator for Redshift in mid-2020 – giving you the ability to scale storage and compute separately and get 10x better query performance than other cloud databases.
7. Build data lakes from data silos to make machine learning possible.
Companies that thrive adapt to use new tools to meet new technology challenges. Organisations that have lots of data silos will find machine learning painful, expensive and slow. Organisations need to build data lakes from data silos – accepting data from multiple sources. This can be done with S3.
8. Get new tools to make machine learning more accessible, faster, and easier.
More people are choosing AWS for machine learning, and so they’ve launched the first fully-integrated environment (IDE) for machine learning to make every step easier – called SageMaker Studio. It stores everything in one place and covers model training, notebooks, managing different iterations and even automated ways to detect concept drift.
In addition to launching SageMaker Studio, AWS has made 50+ innovations in SageMaker alone. They know that developers and data scientists use multiple frameworks for machine learning but don’t have time to transfer algorithms from one to another so with SageMaker have created one place to run the most popular machine learning frameworks. In addition, AWS has created AutoML – giving you full control and visibility over automating machine learning models. It allows you to take automatically created models and evolve them. You can make trade-offs yourself, comparing each model’s accuracy vs latency to select the final model you want to use.
9. Harness machine learning and automation to make every-day tasks easier.
There are plenty of innovations on the day-to-day activities too. AWS launched CodeGuru – giving the ability to automate code reviews so you don’t have to do it manually. CodeGuru also finds the most expensive lines of code, and automatically optimises them. They’ve also launched Kendra – which uses machine learning to create your own enterprise search engine – so you can use natural language to get accurate, specific answers from anywhere in your data.
10. Transform your capability with 5G.
AWS will offer 5G edge cloud computing by teaming up with Verizon in the US to give single-digit latency on mobile and connected devices. They will be teaming up with Vodafone in the UK. This announcement is an exciting development that is one to watch. It will open up the possibilities from real-time data, having great potential to make an impact on creating smart cities and real-time traffic routing.
OK, so there’s one more exciting announcement – blink and you would have missed it
Andy Jassy, CEO of AWS, also briefly mentioned in his keynote speech that AWS have launched multi-cast IP – making it easy for customers to deliver a single stream of data to many users simultaneously. This has huge potential – in the financial services industry for example – where many workloads will benefit.
So there you have it. A busy event with a tonne of exciting announcements. QA is the most experienced authorised training provider for AWS in the UK, offering courses and certifications to keep you up to date with all of the latest technologies and developments. If you are looking to take Andy Jassy’s advice and upskill your existing team in AWS technologies then you can browse our latest programmes at www.qa.com/aws