Last week’s 2016 Amazon Web Services re:Invent was the biggest one yet, with more than 32,000 visitors attending the event across multiple venues throughout Las Vegas. The event included hundreds of breakout sessions covering pretty much every aspect AWS use cases and technologies, Keynotes by AWS CEO Andy Jassy and CTO Werner Vogels, a big sponsors expo, hackathon, labs, and a fun party with a world-class DJ.

Multiple themes were apparent at re:Invent this year; from the event’s key buzzword – “Serverless” – to Big Data, Artificial Intelligence, Containers, and Alexa. While some of the very impressive list of announcements could be considered incremental (new instance types, dev-ops and automations features, management tools, etc.), there were a few announcements and themes in particular that caught my attention from the perspective of their potential impact on early stage startups, which I’d like to share.

1. Serverless: Scaling is easier than ever before

If you think of the evolution of computing, you’d probably start by recalling the physical servers we used to have at our datacenters, then VMs in the datacenter, then VMs in the cloud, and then containers in the cloud. As IT evolved, companies still needed to administer VMs, plan for capacity and utilization, think about fault tolerance, and so on, yet no one wants to do this. Serverless comes to completely do away with this pattern and offers a “Code runtime as a Service” as opposed to a VMs.

With serverless, you run completely on demand: Your code execution is event driven (imagine code running after receiving new data from an IoT sensor) and continuously scaling, but you only pay for the actual code execution. At its core is AWS Lambda, which is the compute runtime, but multiple AWS services are working together for a completely serverless solution including storage, databases and other components. Use cases vary from back end processing to data processing, to chat bots.

screenshot-from-werner-vogels-presentation-at-aws-reinvent-2016Screenshot from AWS CTO Werner Vogels’s Keynote in which he talks about serverless.

What does this mean for your startup?
Taking advantage of new serverless capabilities will change how startups deliver (for the better). With built-in readiness for scale you can speed up time to market, dedicate more time to the actual innovation and user experience, remove operational complexities, and increase developers’ productivity. Often when I meet with young startups, I hear things like “we built the product and it’s well adopted and we’re growing our user base, but now we need to completely re-write the server side for scale”. With serverless, you can avoid this, because it allows you to build a product that is infinitely scalable from the get go.

2. AWS joins the AI game with Lex, Polly, and Rekognition

AWS is now entering the AI in the cloud game with multiple newly announced services and APIs. This comes at a time when other big companies like Google, Baidu and Salesforce are announcing new AI APIs and platforms. Amazon is kicking off its Amazon AI suite with 3 new Deep Learning based services including Amazon Polly – a service that turns text into lifelike speech with 47 male and female voices and supports 24 languages, Amazon Rekognition – an image analysis service that can detect objects, scenes, and faces in images, and Amazon Lex (comes from “Alexa”) – a service for building conversational interfaces (i.e. Bots) into any application using voice and text. Now, doing something like sending an image to a bot which then uses lifelike speech to describe the content of the image back to you, can be achieved with a few AWS API calls.

object and scene detection - awsAmazon Rekognition usage example. Source: AWS blog

What does this mean for your startup?
Until recently, many capabilities such as the 3 just announced by AWS required a level of talent that was extremely scarce in order to develop in-house, or to use non-production ready public packages. These technologies are now becoming accessible to the masses, which means that they would no longer be a differentiating edge for your product, nor would you be able to outdo in-house what the well-funded teams at the huge cloud platforms are already providing. You’d be better off accepting this, and focusing on everything you could build on top of these core technologies to make your product a winner.

Example of banking information delivered through an Amazon Lex chatbotExample of banking information delivered through an Amazon Lex chatbot (serverless of course). Source: AWS blog

3. AWS is fully Container-ready

Containers are already in full swing (you can read about it more in our blog post about our investment in Codefresh – a Docker-native CI/CD platform). On AWS there are multiple ways to implement microservices using Docker, from tailor made solutions you build yourself with open source tools, to Docker Datacenter, and all the way to the ‘out of the box’ Amazon EC2 Container Service (ECS). Of the options available for managing container on AWS, ECS is the solution that fully integrates with the other AWS services including the fine-grained security modeling.

What does this mean for your startup?
No more excuses! AWS is fully production-ready for microservices built on top of Docker containers. Startups can adopt containers on the cloud and use it to get to a “continuous everything” quickly and efficiently. Tying this to the Serverlesss discussion above, real-world implementations will most likely benefit from microservices as well as serverless. In the end, it will help you get customer feedback faster, deliver a better solution more efficiently, grow your business, and place you in a better position to reach the milestones of your next financing round targets.

the continuous everything... nirvanaPresentation screenshot from the session on Development Workflow with Docker and Amazon ECS

4. Amazon Athena: Serverless Analytics over Raw Data

Amazon Athena is a new service that makes it super easy to analyze data in Amazon S3 using standard SQL. This is (obviously) a serverless solution so there is nothing to set up, and there’s no need to do an ETL. You are basically getting the experience of running queries directly on your data files in their raw format. Easy, right?

What does this mean for your startup?
Amazon Athena now gives startups access to powerful data crunching earlier on in their lifecycle. Even if your team is lacking in data ‘heavy lifting’ professionals and you are not ready to build a more complete big data and analytics architecture for your cloud app, you can take advantage of Athena to gain insights quickly with only SQL, in a Pay Per Query model.

Screenshot from AWS re:Invent 2016 sessio on AWS AthenaPresentation screenshot from the session on best practices and common use cases for Amazon Athena

5. AWS in the IoT device with AWS Greengrass

Servers are moving to the Cloud but most IoT devices are, and will, stay on prem. This requires better alignment between what’s in an IoT device and what’s in the cloud. AWS Greengrass takes AWS beyond the cloud and enables local compute, messaging, and data caching for connected devices. To quote AWS: “AWS Greengrass seamlessly extends AWS to devices so they can act locally on the data they generate, while still using the cloud for management, analytics, and durable storage”.

What does this mean for your startup?
If you are a startup building connected devices, platforms such as AWS Greengrass might help you create next generation connected devices that are far more capable at the edge, overcoming edge challenges such as round-trip latency, connectivity issues, and bandwidth cost. These are very early days for AWS Greengrass but it may fit well if your cloud side is AWS and you want to make the most out of a solution that’s basically built on top of AWS IoT and AWS Lambda, not to mention benefit from the security features of AWS Greengrass, which addresses a major problem for IoT (security).

In conclusion…

AWS re:Invent 2016 reminded everyone just how cutting-edge Amazon are with their AWS offerings, and that AWS are still worthy of their top-right position in Gartner’s Magic Quadrant for Cloud. As new improvements and services emerge each year, it becomes easier for new startups to benefit from capabilities that only a handful of hard-to-find talent could handle just a couple of years earlier.

Carmel Ventures portfolio companies exhibiting at this year's AWS re:Invent 2016 (Cloudyn, Codefresh and Redis Labs)Some of Carmel’s portfolio companies exhibiting at AWS re:Invent 2016: Cloudyn, Codefresh and Redis Labs (click to enlarge image)

Relevant links to check out:
bullet-pink-20x10The complete roundup of all the product announcements from this year’s AWS re:Invent
bullet-pink-20x10AWS re:Invent 2016 Keynote by Andy Jassy, CEO of Amazon Web Services
bullet-pink-20x10AWS re:Invent 2016 Keynote by Werner Vogels, CTO
bullet-pink-20x10List of all of the AWS re:Invent 2016 Breakout Sessions


More posts by Ran Levitzky:
bullet-pink-20x105 Key Security Takeaways from RSA Conference 2016
bullet-pink-20x10Carmel Ventures invests in Cloudyn, the startup empowering Enterprises to embrace the hybrid cloud
bullet-pink-20x10Why startups must protect access to their cloud services from Day 1 (and how to do it in 3 easy steps)