AWS re:Invent: Building AI-driven businesses for the future
AWS re:Invent (December 1st – 5th 2025, Las Vegas) has always been a barometer for where cloud technology is heading, but this year there was a consistent and resounding message – AI is no longer an add-on. It is becoming the foundation of how modern businesses operate.
December 2025 marked my first time attending re:Invent. I’ll confess, it is an awe-inspiring event. For Leighton – as a business that specialises in AWS cloud enablement and digital product engineering – the event offered a powerful perspective on both the scale of AWS’ ambition and the opportunity opening up for organisations willing to adapt. What stood out wasn’t just the pace of innovation, but the intent behind it. AWS is no longer positioning AI as an emerging capability, but as a core pillar of its platform, operating model and customer strategy.
Across keynotes and product announcements, it was regularly reinforced that an AI-first mindset is rapidly becoming a strategic necessity. From autonomous AI agents moving from hype to real-world impact, to custom models becoming accessible to a far broader set of organisations, it was clear that AI has moved well beyond experimentation. AWS is embedding AI deeply into both how it runs its own business and how it enables customers.
For organisations watching from the sidelines, this signals a turning point. The gap between legacy platforms and AI-ready architectures is widening, and the decisions businesses make now around modernisation will define how competitive, resilient and adaptable they can be in the years ahead.
AWS made a very clear statement at re:Invent – AI is now central to its entire business strategy. This wasn’t just about showcasing clever demos or experimental tools. The focus was firmly on outcomes and how AI can be embedded into real workloads, governed effectively, and continuously improved.
The underlying message was urgency. The gap between organisations running modern, cloud-native platforms and those still constrained by legacy systems is getting bigger, quickly. Those with the right foundations will move faster, build more, and learn more. Those without them will struggle to keep pace.
One of the strongest themes I took away was how deliberately AWS is positioning itself as the cloud platform for software engineers.
The sheer volume of tools, services and ultimately options being made available points to a future where software engineers will be increasingly supported – and augmented – by AI agents. These agents won’t just assist with snippets of code; they will participate in the full software development lifecycle, effectively part of the team. AWS introduced Kiro – an agentic IDE and CLI – is a great example of this and something our AWS Practice at Leighton is already exploring how to embed in our operations and how we deliver for customers.
As another example, one keynote speaker described teams using agents to progress code out of hours within controlled workflows. That kind of productivity shift feels genuinely game-changing. Engineers and organisations that embrace AI-assisted development will be faster, more predictable, and ultimately more valuable.
A consistent message throughout re:Invent was that modernisation is a prerequisite for AI adoption.
If your platforms, data and applications are not cloud-native, modular and efficient in the face of change, your ability to take advantage of AI will be severely limited. This aligns closely with what we see across our own customer base. Organisations that have invested in modernisation are the ones now able to move quickly with AI.
Put simply, you cannot layer AI effectively on top of outdated architecture. Technical readiness should no longer be viewed as a discretionary investment, but as a survival requirement for organisations that want to remain competitive.
While the opportunity is enormous, AWS was equally clear about the risks.
AI adoption without strong governance, policy controls and observability will create chaos. If you don’t manage AI deliberately, it will spread organically across your organisation often in ways that are unintended, poorly understood and difficult to control. Unchecked experimentation does not scale, it fragments.
The most effective organisations are working to establish guardrails early, start with narrow, well-defined workflows, measuring outcomes rigorously and scaling only once value and risk are clearly understood. This “move fast, but within boundaries” approach allows teams to innovate without losing control and was a recurring theme across customer case studies.
One of the more counterintuitive insights from re:Invent was that greater engineering efficiency is likely to increase demand, not reduce it.
As building software becomes quicker and more predictable, organisations will want to do more with it meaning greater demand for skilled engineers who can work with these tools effectively. However, there is also a looming skills challenge. Several speakers highlighted that it is effectively impossible to hire people with deep, real-world experience in AI-first engineering today simply because the technology is moving so fast.
That creates a clear advantage for organisations willing to invest now because they will develop the insight and the skill set that everyone else wants later. A progressive approach to allowing engineering teams to methodically explore how to work with these tools will help organisations to avoid playing catch-up later down the line.
The advent of AI tools to support our teams is not just a technical shift. It’s an operational and organisational one that affects how teams work, how decisions are made, and how value is delivered. So where should the focus be right now?
· Treat modernisation as non-negotiable - Get your technology estate into a position where AI adoption is genuinely feasible.
· Design AI foundations before scaling - Define policies, guardrails, operating models and data strategies early.
· Start small and prove value - Focus on a single workflow, learn quickly, measure outcomes, then expand.
· Upskill deliberately - Build AI capability inside your organisation rather than waiting for it to appear in the market.
AWS re:Invent reinforced for me that this is a once-in-a-generation opportunity for organisations, for engineers, and for business leaders willing to move decisively. But, with opportunity comes risk. Organisations that fail to commercialise AI in ways that make sense for their industry may find themselves left behind.
What excites me most is how closely this aligns with the problems we solve every day with our customers. The conversations sparked at re:Invent are the same ones we’re having on the ground – how to modernise platforms without disrupting day-to-day operations, how to move away from fragmented legacy systems, and how to build foundations that allow AI to deliver real value rather than just promise.
If re:Invent has raised similar questions for you around modernisation, AI readiness or how to build AI into your roadmap so you can achieve practical outcomes – we’d love to talk. You can contact the team here or drop me a line on LinkedIn.
AWS re:Invent (December 1st – 5th 2025, Las Vegas) has always been a barometer for where cloud technology is heading, but this year there was a consistent and resounding message – AI is no longer an add-on. It is becoming the foundation of how modern businesses operate.
December 2025 marked my first time attending re:Invent. I’ll confess, it is an awe-inspiring event. For Leighton – as a business that specialises in AWS cloud enablement and digital product engineering – the event offered a powerful perspective on both the scale of AWS’ ambition and the opportunity opening up for organisations willing to adapt. What stood out wasn’t just the pace of innovation, but the intent behind it. AWS is no longer positioning AI as an emerging capability, but as a core pillar of its platform, operating model and customer strategy.
Across keynotes and product announcements, it was regularly reinforced that an AI-first mindset is rapidly becoming a strategic necessity. From autonomous AI agents moving from hype to real-world impact, to custom models becoming accessible to a far broader set of organisations, it was clear that AI has moved well beyond experimentation. AWS is embedding AI deeply into both how it runs its own business and how it enables customers.
For organisations watching from the sidelines, this signals a turning point. The gap between legacy platforms and AI-ready architectures is widening, and the decisions businesses make now around modernisation will define how competitive, resilient and adaptable they can be in the years ahead.
AWS made a very clear statement at re:Invent – AI is now central to its entire business strategy. This wasn’t just about showcasing clever demos or experimental tools. The focus was firmly on outcomes and how AI can be embedded into real workloads, governed effectively, and continuously improved.
The underlying message was urgency. The gap between organisations running modern, cloud-native platforms and those still constrained by legacy systems is getting bigger, quickly. Those with the right foundations will move faster, build more, and learn more. Those without them will struggle to keep pace.
One of the strongest themes I took away was how deliberately AWS is positioning itself as the cloud platform for software engineers.
The sheer volume of tools, services and ultimately options being made available points to a future where software engineers will be increasingly supported – and augmented – by AI agents. These agents won’t just assist with snippets of code; they will participate in the full software development lifecycle, effectively part of the team. AWS introduced Kiro – an agentic IDE and CLI – is a great example of this and something our AWS Practice at Leighton is already exploring how to embed in our operations and how we deliver for customers.
As another example, one keynote speaker described teams using agents to progress code out of hours within controlled workflows. That kind of productivity shift feels genuinely game-changing. Engineers and organisations that embrace AI-assisted development will be faster, more predictable, and ultimately more valuable.
A consistent message throughout re:Invent was that modernisation is a prerequisite for AI adoption.
If your platforms, data and applications are not cloud-native, modular and efficient in the face of change, your ability to take advantage of AI will be severely limited. This aligns closely with what we see across our own customer base. Organisations that have invested in modernisation are the ones now able to move quickly with AI.
Put simply, you cannot layer AI effectively on top of outdated architecture. Technical readiness should no longer be viewed as a discretionary investment, but as a survival requirement for organisations that want to remain competitive.
While the opportunity is enormous, AWS was equally clear about the risks.
AI adoption without strong governance, policy controls and observability will create chaos. If you don’t manage AI deliberately, it will spread organically across your organisation often in ways that are unintended, poorly understood and difficult to control. Unchecked experimentation does not scale, it fragments.
The most effective organisations are working to establish guardrails early, start with narrow, well-defined workflows, measuring outcomes rigorously and scaling only once value and risk are clearly understood. This “move fast, but within boundaries” approach allows teams to innovate without losing control and was a recurring theme across customer case studies.
One of the more counterintuitive insights from re:Invent was that greater engineering efficiency is likely to increase demand, not reduce it.
As building software becomes quicker and more predictable, organisations will want to do more with it meaning greater demand for skilled engineers who can work with these tools effectively. However, there is also a looming skills challenge. Several speakers highlighted that it is effectively impossible to hire people with deep, real-world experience in AI-first engineering today simply because the technology is moving so fast.
That creates a clear advantage for organisations willing to invest now because they will develop the insight and the skill set that everyone else wants later. A progressive approach to allowing engineering teams to methodically explore how to work with these tools will help organisations to avoid playing catch-up later down the line.
The advent of AI tools to support our teams is not just a technical shift. It’s an operational and organisational one that affects how teams work, how decisions are made, and how value is delivered. So where should the focus be right now?
· Treat modernisation as non-negotiable - Get your technology estate into a position where AI adoption is genuinely feasible.
· Design AI foundations before scaling - Define policies, guardrails, operating models and data strategies early.
· Start small and prove value - Focus on a single workflow, learn quickly, measure outcomes, then expand.
· Upskill deliberately - Build AI capability inside your organisation rather than waiting for it to appear in the market.
AWS re:Invent reinforced for me that this is a once-in-a-generation opportunity for organisations, for engineers, and for business leaders willing to move decisively. But, with opportunity comes risk. Organisations that fail to commercialise AI in ways that make sense for their industry may find themselves left behind.
What excites me most is how closely this aligns with the problems we solve every day with our customers. The conversations sparked at re:Invent are the same ones we’re having on the ground – how to modernise platforms without disrupting day-to-day operations, how to move away from fragmented legacy systems, and how to build foundations that allow AI to deliver real value rather than just promise.
If re:Invent has raised similar questions for you around modernisation, AI readiness or how to build AI into your roadmap so you can achieve practical outcomes – we’d love to talk. You can contact the team here or drop me a line on LinkedIn.
AWS re:Invent (December 1st – 5th 2025, Las Vegas) has always been a barometer for where cloud technology is heading, but this year there was a consistent and resounding message – AI is no longer an add-on. It is becoming the foundation of how modern businesses operate.
December 2025 marked my first time attending re:Invent. I’ll confess, it is an awe-inspiring event. For Leighton – as a business that specialises in AWS cloud enablement and digital product engineering – the event offered a powerful perspective on both the scale of AWS’ ambition and the opportunity opening up for organisations willing to adapt. What stood out wasn’t just the pace of innovation, but the intent behind it. AWS is no longer positioning AI as an emerging capability, but as a core pillar of its platform, operating model and customer strategy.
Across keynotes and product announcements, it was regularly reinforced that an AI-first mindset is rapidly becoming a strategic necessity. From autonomous AI agents moving from hype to real-world impact, to custom models becoming accessible to a far broader set of organisations, it was clear that AI has moved well beyond experimentation. AWS is embedding AI deeply into both how it runs its own business and how it enables customers.
For organisations watching from the sidelines, this signals a turning point. The gap between legacy platforms and AI-ready architectures is widening, and the decisions businesses make now around modernisation will define how competitive, resilient and adaptable they can be in the years ahead.
AWS made a very clear statement at re:Invent – AI is now central to its entire business strategy. This wasn’t just about showcasing clever demos or experimental tools. The focus was firmly on outcomes and how AI can be embedded into real workloads, governed effectively, and continuously improved.
The underlying message was urgency. The gap between organisations running modern, cloud-native platforms and those still constrained by legacy systems is getting bigger, quickly. Those with the right foundations will move faster, build more, and learn more. Those without them will struggle to keep pace.
One of the strongest themes I took away was how deliberately AWS is positioning itself as the cloud platform for software engineers.
The sheer volume of tools, services and ultimately options being made available points to a future where software engineers will be increasingly supported – and augmented – by AI agents. These agents won’t just assist with snippets of code; they will participate in the full software development lifecycle, effectively part of the team. AWS introduced Kiro – an agentic IDE and CLI – is a great example of this and something our AWS Practice at Leighton is already exploring how to embed in our operations and how we deliver for customers.
As another example, one keynote speaker described teams using agents to progress code out of hours within controlled workflows. That kind of productivity shift feels genuinely game-changing. Engineers and organisations that embrace AI-assisted development will be faster, more predictable, and ultimately more valuable.
A consistent message throughout re:Invent was that modernisation is a prerequisite for AI adoption.
If your platforms, data and applications are not cloud-native, modular and efficient in the face of change, your ability to take advantage of AI will be severely limited. This aligns closely with what we see across our own customer base. Organisations that have invested in modernisation are the ones now able to move quickly with AI.
Put simply, you cannot layer AI effectively on top of outdated architecture. Technical readiness should no longer be viewed as a discretionary investment, but as a survival requirement for organisations that want to remain competitive.
While the opportunity is enormous, AWS was equally clear about the risks.
AI adoption without strong governance, policy controls and observability will create chaos. If you don’t manage AI deliberately, it will spread organically across your organisation often in ways that are unintended, poorly understood and difficult to control. Unchecked experimentation does not scale, it fragments.
The most effective organisations are working to establish guardrails early, start with narrow, well-defined workflows, measuring outcomes rigorously and scaling only once value and risk are clearly understood. This “move fast, but within boundaries” approach allows teams to innovate without losing control and was a recurring theme across customer case studies.
One of the more counterintuitive insights from re:Invent was that greater engineering efficiency is likely to increase demand, not reduce it.
As building software becomes quicker and more predictable, organisations will want to do more with it meaning greater demand for skilled engineers who can work with these tools effectively. However, there is also a looming skills challenge. Several speakers highlighted that it is effectively impossible to hire people with deep, real-world experience in AI-first engineering today simply because the technology is moving so fast.
That creates a clear advantage for organisations willing to invest now because they will develop the insight and the skill set that everyone else wants later. A progressive approach to allowing engineering teams to methodically explore how to work with these tools will help organisations to avoid playing catch-up later down the line.
The advent of AI tools to support our teams is not just a technical shift. It’s an operational and organisational one that affects how teams work, how decisions are made, and how value is delivered. So where should the focus be right now?
· Treat modernisation as non-negotiable - Get your technology estate into a position where AI adoption is genuinely feasible.
· Design AI foundations before scaling - Define policies, guardrails, operating models and data strategies early.
· Start small and prove value - Focus on a single workflow, learn quickly, measure outcomes, then expand.
· Upskill deliberately - Build AI capability inside your organisation rather than waiting for it to appear in the market.
AWS re:Invent reinforced for me that this is a once-in-a-generation opportunity for organisations, for engineers, and for business leaders willing to move decisively. But, with opportunity comes risk. Organisations that fail to commercialise AI in ways that make sense for their industry may find themselves left behind.
What excites me most is how closely this aligns with the problems we solve every day with our customers. The conversations sparked at re:Invent are the same ones we’re having on the ground – how to modernise platforms without disrupting day-to-day operations, how to move away from fragmented legacy systems, and how to build foundations that allow AI to deliver real value rather than just promise.
If re:Invent has raised similar questions for you around modernisation, AI readiness or how to build AI into your roadmap so you can achieve practical outcomes – we’d love to talk. You can contact the team here or drop me a line on LinkedIn.
AWS re:Invent (December 1st – 5th 2025, Las Vegas) has always been a barometer for where cloud technology is heading, but this year there was a consistent and resounding message – AI is no longer an add-on. It is becoming the foundation of how modern businesses operate.
December 2025 marked my first time attending re:Invent. I’ll confess, it is an awe-inspiring event. For Leighton – as a business that specialises in AWS cloud enablement and digital product engineering – the event offered a powerful perspective on both the scale of AWS’ ambition and the opportunity opening up for organisations willing to adapt. What stood out wasn’t just the pace of innovation, but the intent behind it. AWS is no longer positioning AI as an emerging capability, but as a core pillar of its platform, operating model and customer strategy.
Across keynotes and product announcements, it was regularly reinforced that an AI-first mindset is rapidly becoming a strategic necessity. From autonomous AI agents moving from hype to real-world impact, to custom models becoming accessible to a far broader set of organisations, it was clear that AI has moved well beyond experimentation. AWS is embedding AI deeply into both how it runs its own business and how it enables customers.
For organisations watching from the sidelines, this signals a turning point. The gap between legacy platforms and AI-ready architectures is widening, and the decisions businesses make now around modernisation will define how competitive, resilient and adaptable they can be in the years ahead.
AWS made a very clear statement at re:Invent – AI is now central to its entire business strategy. This wasn’t just about showcasing clever demos or experimental tools. The focus was firmly on outcomes and how AI can be embedded into real workloads, governed effectively, and continuously improved.
The underlying message was urgency. The gap between organisations running modern, cloud-native platforms and those still constrained by legacy systems is getting bigger, quickly. Those with the right foundations will move faster, build more, and learn more. Those without them will struggle to keep pace.
One of the strongest themes I took away was how deliberately AWS is positioning itself as the cloud platform for software engineers.
The sheer volume of tools, services and ultimately options being made available points to a future where software engineers will be increasingly supported – and augmented – by AI agents. These agents won’t just assist with snippets of code; they will participate in the full software development lifecycle, effectively part of the team. AWS introduced Kiro – an agentic IDE and CLI – is a great example of this and something our AWS Practice at Leighton is already exploring how to embed in our operations and how we deliver for customers.
As another example, one keynote speaker described teams using agents to progress code out of hours within controlled workflows. That kind of productivity shift feels genuinely game-changing. Engineers and organisations that embrace AI-assisted development will be faster, more predictable, and ultimately more valuable.
A consistent message throughout re:Invent was that modernisation is a prerequisite for AI adoption.
If your platforms, data and applications are not cloud-native, modular and efficient in the face of change, your ability to take advantage of AI will be severely limited. This aligns closely with what we see across our own customer base. Organisations that have invested in modernisation are the ones now able to move quickly with AI.
Put simply, you cannot layer AI effectively on top of outdated architecture. Technical readiness should no longer be viewed as a discretionary investment, but as a survival requirement for organisations that want to remain competitive.
While the opportunity is enormous, AWS was equally clear about the risks.
AI adoption without strong governance, policy controls and observability will create chaos. If you don’t manage AI deliberately, it will spread organically across your organisation often in ways that are unintended, poorly understood and difficult to control. Unchecked experimentation does not scale, it fragments.
The most effective organisations are working to establish guardrails early, start with narrow, well-defined workflows, measuring outcomes rigorously and scaling only once value and risk are clearly understood. This “move fast, but within boundaries” approach allows teams to innovate without losing control and was a recurring theme across customer case studies.
One of the more counterintuitive insights from re:Invent was that greater engineering efficiency is likely to increase demand, not reduce it.
As building software becomes quicker and more predictable, organisations will want to do more with it meaning greater demand for skilled engineers who can work with these tools effectively. However, there is also a looming skills challenge. Several speakers highlighted that it is effectively impossible to hire people with deep, real-world experience in AI-first engineering today simply because the technology is moving so fast.
That creates a clear advantage for organisations willing to invest now because they will develop the insight and the skill set that everyone else wants later. A progressive approach to allowing engineering teams to methodically explore how to work with these tools will help organisations to avoid playing catch-up later down the line.
The advent of AI tools to support our teams is not just a technical shift. It’s an operational and organisational one that affects how teams work, how decisions are made, and how value is delivered. So where should the focus be right now?
· Treat modernisation as non-negotiable - Get your technology estate into a position where AI adoption is genuinely feasible.
· Design AI foundations before scaling - Define policies, guardrails, operating models and data strategies early.
· Start small and prove value - Focus on a single workflow, learn quickly, measure outcomes, then expand.
· Upskill deliberately - Build AI capability inside your organisation rather than waiting for it to appear in the market.
AWS re:Invent reinforced for me that this is a once-in-a-generation opportunity for organisations, for engineers, and for business leaders willing to move decisively. But, with opportunity comes risk. Organisations that fail to commercialise AI in ways that make sense for their industry may find themselves left behind.
What excites me most is how closely this aligns with the problems we solve every day with our customers. The conversations sparked at re:Invent are the same ones we’re having on the ground – how to modernise platforms without disrupting day-to-day operations, how to move away from fragmented legacy systems, and how to build foundations that allow AI to deliver real value rather than just promise.
If re:Invent has raised similar questions for you around modernisation, AI readiness or how to build AI into your roadmap so you can achieve practical outcomes – we’d love to talk. You can contact the team here or drop me a line on LinkedIn.
