Written by Dale Gates
Once again I found myself travelling to the capital for what has become a yearly pilgrimage to the AWS London Summit hosted at the Excel Centre. The train journey from Newcastle gave me ample time to ponder what would be the talk of the year and the strategic direction AWS would unveil. I had an inkling it would continue to revolve around AI and LLMs—after all, the last two years had seen the likes of generative AI continue to dominate discussions—but would there be something new?
Keynote Highlights
The opening keynote confirmed my expectations—AWS continues to bet heavily on AI, but with several notable announcements beyond the expected LLM updates. The UKI Lead’s commitment to not just carbon neutrality but pushing toward carbon negative operations by 2030 was well received, showing AWS’s determination to lead on the drive for sustainability.
Infrastructure investment also featured prominently with a strong focus on compute and storage, with AWS working closely with hardware partners to enhance their tailor-made instance types. The advancements in their custom silicon strategy were particularly impressive, with the next generation of AWS Trainium processors designed specifically for AI workloads demonstrating performance gains that could significantly reduce training costs for our customers implementing machine learning solutions.
I was particularly excited to see AWS Deadline Cloud featured prominently in the keynote, getting a 10 minute or so mention. Given Amazon (Not quite AWS but rather the parent company e-shop which satisfies your impulse buys) have expanded their portfolio and are now firmly embedded within the film and TV industry, it should really be unsurprising that they have started to utilise their own product.
As a cloud-based version of the well-established Thinkbox Deadline 10, this solution is designed to revolutionise how visual effects, animation, and other compute-intensive media production workloads are rendered in the cloud. It represents a significant opportunity for several of our clients in the creative industries to render with the latest hardware on-demand, without the expensive on-premise hardware renewal dance that occurs every few years.
However, my enthusiasm was somewhat dampened when I discovered a notable absence of breakout workshops or deeper technical sessions on Deadline Cloud’s implementation and optimisation. I’m unsure if this is just due to the relative immaturity of the product (AWS Deadline Cloud not the aforementioned On-Premise offering which is quite the opposite), but I guess we need to watch this space.
This to me represents a clear opportunity gap that we at Expert Thinking are well-positioned to fill, given the recent direction of the AWS BU and our close partnership with the visual compute team at AWS.
AI
As predicted, AI continues to dominate discussions at the Summit, but this year the focus has shifted noticeably from theoretical capabilities to genuine business impact across industries. The advancements in healthcare are particularly impressive, with demonstrations showing how image processing combined with AI is enabling earlier and more accurate cancer detection – to go off on a related tangent an article that caught my eye had one model recently having an almost 100% detection rate during testing Nearly 100% of cancer identified by new AI, easily outperforming doctors
The financial sector has also begun realising substantial benefits. NatWest showcased their implementation of AWS’ AI services, which they credited with preventing approximately £300 million in fraudulent transactions over the past year. Their representatives described how the system’s ability to analyse patterns across vast datasets has dramatically improved their ability to detect and prevent financial crimes before they impact customers.
These case studies present compelling evidence that AI has moved beyond the hype cycle into delivering measurable business value and real world positive change — a narrative that will resonate strongly with our customers who may have been initially hesitant to invest in these technologies.
Partner Ecosystem
The partner exhibition hall featured the usual mix of established players and ambitious newcomers. I couldn’t help but smile at the number of new faces that had clearly shoehorned “AI” somewhere into their brand names—a telling indicator of where the industry believes the future lies.
Anthropic maintained a substantial presence, showcasing Claude’s capabilities across various enterprise applications – with Claude playing classic Pokemon in one of the demo areas and giving us the real content that we’re here for.
Perhaps most telling was NVIDIA’s pavilion—larger than many of the other established partners. This expanded footprint is unsurprising given the number of GPU instances NVIDIA will be supplying to AWS data centres and the heavy reliance of GPU utilisation for ML and LLM workloads . Their representatives were particularly keen to discuss optimisation strategies for enterprise AI workloads.
Workshops and Talks
As previously mentioned despite Deadline Cloud’s prominent mention in the keynote, I was surprised to find absolutely no workshops or technical sessions dedicated to this potentially transformative service. This absence struck me as a prime opportunity for Expert Thinking to establish thought leadership in this space—we could potentially be up there presenting next year or even later this year at re-invent, given our experience in this relatively niche area.
In contrast, there was a noticeable increase in sessions focused on Nova, AWS’s answer to image generation tools like DALL-E. The demonstrations were genuinely impressive, showcasing capabilities for transforming still images into short videos, removing and changing backgrounds with remarkable precision, and generating completely new video content from text prompts. These capabilities could dramatically reduce content production costs for marketers or even lower the bar for smaller businesses such as E shops to give their products a bit more “pizzaz”.
And as already mentioned – AI took up the vast majority of the sessions with a quick search off “AI” in the events returning 234 sessions out of 279.
The Fun(er) Section
As always, AWS and their partners created the engaging “Fun(er)” experiences throughout one of the exhibition halls.
Bundesliga’s free kick challenge (was more a penalty but nevermind) drew consistent crowds, using sophisticated camera systems for ball detection and shot analysis, complete with AI-generated commentary that was eerily similar to professional broadcasters. By the end of the day, I ranked about 20th overall with a put your laces through it approach then opting to change to the technique to a toe punt which still wasn’t quite enough to secure a leaderboard position.
Formula 1 maintained its prominent presence, with a tyre changing challenge that had attendees competing for the fastest pit stop times. My Friends who also attended the event managed an impressive 3.2 seconds, topping the leaderboards for much of the day. Mercedes contract offers are surely inbound and I was rather impressed by their unexpected dexterity.
Finally a travellator experience in which I was fully expecting Wolf from gladiators to make an appearance at the summit. Once again the same friend who excelled at changing tyres pumped an impressive 3 second time to reach the peak (Yes – He is THAT friend) whilst a slow 10 seconds from myself
These gamified experiences however do more than entertain—they showcase the practical applications of AWS services and technologies in areas such as sports analytics and real-time data processing with reference architectures being showcased at each of the experiences.
Final Word
As we travelled to one of the many afterparties, bags filled with branded swag, notebook (from said swag) filled with notes of the day, I reflected on this vs last year. While generative AI continues to dominate the conversation, we’re finally seeing the shift from technological novelty to business necessity—complete with compelling ROI metrics and case studies that will help us justify investment to our customers.
The opportunities for Expert Thinking are substantial, particularly in underserved areas like AWS Deadline Cloud but also in helping organisations move beyond AI experimentation toward enterprise-scale deployment. Our position as trusted advisors becomes increasingly valuable as the technology landscape continues to grow more complex.
The AWS London Summit may have ended, but the real work—translating these insights into tangible value for our customers—is just beginning. And that’s precisely where we as Principals earn our keep: bridging the gap between the new cutting edge dazzling technologies and practical business outcomes.