AWS Summit London is a free cloud event at ExCeL London. Starting from 8:00 am you could grab a schedule, pick up your badge and wander an expo floor divided into zones like AI/ML, databases, security, networking, and serverless.

The day kicks off with a one-hour keynote featuring new AWS services and customer success stories, then you choose from sessions, chalk talks, workshops, or live demos.

The expo had partner and startup booths where you can try out data pipeline frameworks, see how companies are using their own services, edge compute appliances and more. Some engineers and founders are at hand to run live demos and dive into how their products solve real problems.

AWS experts and partners are everywhere too, ready to answer questions.

My highlights of the day:

I’ll give a breakdown of some of my favourite highlights. Some may be very technical but I hope to break it down so you understand each aspect and how AWS uses their services.

Augmented reality gaming experience:

This demo turned a regular tabletop into an AR chess game using a full AWS stack. Amazon Bedrock generated AI-driven chess opponents and move suggestions on the go. Every piece’s position and player stats were stored in DynamoDB, so the game state stayed consistent across sessions. Lambda functions ran the game logic, validating moves, checking for checkmate and updating scores!

AppSync handled real-time syncing, so both players saw each other’s moves instantly. Amazon Cognito managed user sign-in and authentication, keeping games private and secure.

All the 3D models and textures sat in S3 and were delivered globally through CloudFront, so assets loaded fast with minimal lag. It was a slick proof of how you can blend physical boards with cloud-powered AR to create a seamless, interactive experience. The image below shows the architecture of this game! Yes I know, how cool!!

Hosting DeepSeek-R1 on Amazon EKS

At the AWS Summit London, this booth is showing how they’ve deployed an AI model called DeepSeek-R1 using Amazon EKS (Elastic Kubernetes Service). EKS is basically a managed version of Kubernetes that AWS takes care of for you. It lets you run containers (like Docker apps) at scale without needing to manage all the complicated parts like networking, cluster upgrades, and security by yourself.

This architectural diagram explaining how DeepSeek-R1 is deployed on Amazon EKS (Elastic Kubernetes Service). Amazon EKS Cluster is the central Kubernetes control plane that manages your containerized application.

On the big screen, the architecture diagram also splits the responsibilities. AWS manages the control parts of the cluster (like the brain of Kubernetes), while you (the customer) manage things like your own load balancers, volumes for data, and the actual app you’re running.

In this case, DeepSeek-R1 (an AI language model similar to ChatGPT) is being run inside the EKS cluster. That means it’s running on containers, spread across powerful EC2 machines, possibly with GPUs and connected to AWS services like storage, networking, and security. This setup gives teams the ability to deploy advanced AI apps quickly and reliably, scaling up or down as needed and using the power of the cloud to handle big workloads without needing to build their own data centre.

Edge / Networking / Security Observability

It’s all about how AWS helps teams monitor and secure their infrastructure, especially when it’s distributed across cloud environments, edge locations and hybrid networks. The main focus here is AWS Security Lake. This is a service that helps you centralise and analyse all your security-related data in one place. It pulls logs and alerts from various sources like:

• AWS CloudTrail: records API activity (who did what and when).

• VPC Flow Logs: captures network traffic going in/out of your AWS environment.

• DNS Logs, Security agents, GuardDuty: for threat detection and behaviour tracking.

All this data gets stored in Amazon Security Lake, which is built on top of Amazon S3 for long-term, cost-effective storage. It follows the “write once, read many” model, meaning the data can be reused by many tools without duplication.

Instead of manually gathering logs from different services and trying to piece them together, Security Lake automatically formats everything. That means security teams (InfoSec, SIEM tools, threat hunting apps, etc.) can run analytics, detect threats faster and automate responses.

Sports Zone Freekick Challenge

This shows how AWS services are used to track, analyse, and display live football match data such as ball movement, player positioning and statistics at the summit’s on-site pitch.

Let’s break it down simply into 2 sections, AWS services and real life input:

1. This starts with real-world input: Ball sensors, player tracking, cameras, and even NVIDIA Jetson devices (used for edge computing). These collect real-time stats like speed, location, ball possession and more.

2. This data gets passed into various AWS services:

• AWS AppSync - manages real-time communication between front-end apps and backend. Amazon DynamoDB - stores game state, player stats, scores. AWS Lambda - runs backend logic like calculating leaderboards.

• Amazon ECS - hosts microservices or containers for processing (e.g. game engine or match controller).

• Amazon Bedrock - used here for generating live commentary using GenAI models!

• Guardrail - ensures the generated commentary is safe, appropriate, and on-topic.

• Amazon Cognito – handles player or user authentication for apps.

• Attendee web app, operator station, scoreboard, leaderboard display which is all powered by this backend.

• Feeds show live match commentary, stats, and visuals.

• Amazon S3 + CloudFront, stores and delivers media/content quickly

• AWS WAF - protects the apps from cyber threats.

• Amazon SageMaker - trains ML models (e.g. ball tracking or player predictions).

This may look overwhelming but don’t worry, all of these services work together to provide a live football game powered by AI and AWS. Sensors and cameras feed real-time data into AWS services, which then analyses the game and updates a live scoreboard. Furthermore, it can generate commentary and stream it all securely through the cloud.

It’s a cool showcase of how AWS can bring together edge computing, AI/ML and cloud infrastructure for real-time sports tech.

Amazon Q Developer

Amazon Q Developer is a GenAI-powered assistant for developers built to help you write, understand, and troubleshoot code, especially for Infrastructure as Code (IaC) like Terraform or AWS CloudFormation. It can explain what code is doing before you make changes.

Write Terraform/CloudFormation code and describe what you want in plain English. It writes the infrastructure code for you.

Some of the things Amazon Q developer can do:

• Scan your code for errors or security issues: Very useful for spotting problems early.

• Troubleshoot errors: If something breaks during deployment, Q can help fix it.

• Chat about resources: You can ask, “What does this Lambda function do?” or “Why is this bucket public?”

• Create pull requests & README files: It automates documentation and GitHub tasks to save time.

If you’re working with AWS infrastructure, Terraform, or DevOps pipelines, Amazon Q Developer is like having a super-smart assistant inside your IDE (like VS Code). It helps you work faster, reduce errors and learn on the go especially if you’re still growing your IaC or cloud skills.

Ask an Architect:

At the “Ask an Architect” station, I grabbed a few minutes with a Technical Account Manager who laid out exactly what he does day-to-day. Everything from reviewing customers’ cloud architectures and identifying cost or performance bottlenecks, to coordinating incident response when things go sideways at massive scale.

He walked me through real examples of how AWS steps in early to design resilient, secure systems that can handle millions of requests per minute, and how his team stays plugged into customers’ roadmaps to risk assessment before they hit production.

Hearing about the mix of hands-on troubleshooting, proactive guidance, and long-term strategy that goes into keeping some of the world’s biggest workloads online was eye-opening. It really drove home how AWS doesn’t just sell services, it partners deeply to solve hardcore engineering challenges and left me inspired about the impact you can make in a role like that. Very inspiring.

Networking:

Networking was the real highlight. In casual conversations, I got pointers on AWS tools I’d never tried and invites to join online Slack and Discord communities where experts and community members share real-world tips. I even connected with leading engineers who shared slide decks from their talks. Those quick face-to-face moments led to meetups invites and invites to deep-dive workshops. What starts as a one-day event snowballs into a support network that keeps you learning, troubleshooting and pushing your projects forward long after the summit ends.

Closing thoughts:

If you’re early in your tech career, AWS Summit London is crucial. You’ll get a front-row seat to the latest cloud and AI tools, pick up practical tips you won’t find in tutorials and meet AWS engineers to fellow peers who can help you level up. It’s a way to build your confidence and start networking with people who you can learn from or those that you can teach too. Plus, you’ll leave with clear next steps and resources to keep learning long after the day ends.

Don’t just read about cloud tech… get hands-on, make connections and kick-start your journey.

Don’t forget to grab the AWS Summit app before you go. It’s a lifesaver for planning your day, you can browse the full agenda, bookmark the sessions you actually care about, and get push reminders so you don’t miss your favourite talk. The built-in venue map shows you the exact location of each room and demo booth. It’ll even nudge you when a session you bookmarked is about to start, so you can spend less time juggling paper schedules.

See you at the AWS Summit London 2026!

What is AWS and the AI Practitioner Exam?

You’ve probably used AWS without even knowing it. It’s the cloud platform that runs websites, apps, streaming services and pretty much everything online. AWS has a ton of services and AI/ML is one of its big areas now…

The AI Practitioner exam is basically your “101” in AWS AI/ML—you won’t write a line of code or solve equations, you’ll just show you know which service to use and why.

You’ll get 65 multiple-choice questions in 90 minutes (plus a 30-minute language extension if you need it), and it costs $100.

The test covers core AI/ML ideas like models vs. algorithms, supervised vs. unsupervised learning, and performance metrics; generative AI basics such as tokens, embeddings, transformers, diffusion models, and prompt-engineering; and responsible AI topics like spotting bias, data encryption, IAM roles, and compliance.

You’ll need to know AWS’s AI lineup; Amazon SageMaker

• Amazon Bedrock

• Amazon Q

• Amazon AI / ML Services:

• Concepts of AI (Supervised Learning, Underfitting, Tokens)

All through scenario-based questions (“Your app needs real-time translation, what service do you call?” or “How do you detect model bias?”).

It’s designed for anyone new to AI/ML who just need to understand what AWS’s AI services can do. This may look foreign to you now, but don’t worry, once you study these concepts and services, you’ll know exactly what we’re talking about!

Why did I take this exam?

Generative AI is everywhere right now, news articles, product announcements, even in casual tech chats. I wanted to see how it all actually works under the hood. AWS keeps rolling out new AI services (Bedrock, SageMaker Studio, CodeWhisperer, Titan models, and more), and the fastest way to learn what each one does is to study for this certification. It gives you a great breakdown on how these services work

Plus, with AI moving so fast, I needed a game plan to stay current. The exam topics gave me a checklist: generative AI basics, responsible AI tools, security and governance, and so on. If you want to keep up without getting overwhelmed, prepping for this exam gives you that structure.

Topics You’ll See

The exam breaks down into five areas:

1. AI/ML Fundamentals (20%)
   What’s a model versus an algorithm? Supervised versus unsupervised learning? Common metrics like accuracy, precision, and RMSE.

2. Generative AI Basics (24%)
   How tokens and embeddings work, what transformers and diffusion models are, and the gist of prompt engineering.

3. Foundation Models & Applications (28%)
   Picking or fine-tuning pre-trained models in Bedrock, using retrieval-augmented generation (RAG), and checking output quality.

4. Responsible AI (14%)
   Spotting and reducing bias, understanding fairness and explainability, and auditing models with tools like SageMaker Clarify.

5. Security & Governance (14%)
   Setting up IAM roles, encrypting data with KMS, tracking data lineage, and sticking to rules like GDPR or HIPAA.

Throughout, you’ll match services to real tasks: SageMaker for training and inference, Comprehend for text, Rekognition for images, Lex for chatbots, Polly for speech, Transcribe/Translate for audio, Personalize for recommendations, etc.

Must-Have Study Resources

Don’t worry, you don’t need to hunt all over the web. These cover everything on the syllabus:

• Official exam guide & whitepapers
  Grab the AWS Certified AI Practitioner Exam Guide PDF to see every topic you need, then deep-dive into the AWS AI/ML whitepapers for real-world tips.

• AWS Skill Builder prep plan
  Free, on-demand modules on Skill Builder, each with quick videos and a short quiz so you can test yourself as you go. AWS Certified AI Practitioner

• Udemy courses
  Follow along with instructors like Stephane Maarek to watch live demos in the AWS console. Also check out [NEW] Ultimate AWS Certified AI Practitioner AIF-C01 for up-to-date walkthroughs.

  

• Personal study notes
  Keep a digital doc of bullet-point summaries that map each service and concept back to the exam domains, super handy for last-minute review.

How to study smart

1. Have a plan with a target.

   • Watch videos and read up on fundamentals, generative AI, and foundation models.

   • Hands-on labs and full-length practice exams.

2. Get Hands-On
   Spin up a free-tier AWS account

3. Flag & Review
   In practice tests, flag any question that trips you up. Go back to AWS docs or quick tutorials until it clicks.

4. Focus on Why
   Know why you’d pick one service over another. If you need to pull entities from text, call Comprehend. If you need real-time inference, hit a SageMaker endpoint.

What exam day feels like

• Check-in: At a center, show your ID. Online, you share your screen and scan your room.

• Interface: Simple web UI. Answer, flag if needed, and move on.

• Question Style: Mostly “scenario” questions e.g. “Your app needs real-time translation; which service do you choose?”

• Results: Pass/fail and a score breakdown show up in your AWS account & by email.

Is this certification worth it?

Yes, if you want a solid starting point in cloud AI/ML:

• You get a foundational AWS certification that shows you know the basic AI/ML and generative AI concepts.

• You build a mental map of AWS’s AI services so you can sketch out solutions with confidence.

• If you’re new to AWS, you can follow up with the Cloud Practitioner cert before moving deeper.

• For those already holding a Cloud Practitioner or Associate certification, this lets you jump into AI specialization next.

This certification isn’t for hardcore data science, but it’s the perfect launchpad if you want to learn more about AI/ML on the cloud.

What certification to take next?

After AI Practitioner, pick the path that fits your goals:

1. Cloud Practitioner
   New to AWS? Start here. You’ll learn the basics of AWS architecture, billing, security, and core services.

2. Solutions Architect - Associate
   Want to design real systems? You’ll learn how to build scalable, resilient apps on AWS and plug in AI services where they fit.

3. Data Engineer - Associate
   Into data pipelines? You’ll master ETL with Glue, data lakes on S3, streaming with Kinesis, and prep data for ML.

4. Machine Learning Engineer – Associate
   Ready to dive deeper? You’ll train, tune, and deploy models in SageMaker, set up MLOps pipelines, handle feature engineering, and monitor models in production. Passing this also renews your AI Practitioner cert for another three years.

Pick the one that lines up with where you want to go next—and go for it!!

Subscribe now