AWS Meetup Dublin | 24 March 2026 - Reinvent Recap + More!!

Meetup: AWS UG Dublin

Location details on Meetup page

AGENDA Sponsorship enquiries here Thanks to fourTheorem for sponsoring this event! 6:00 PM - WELCOME Ronan Ó Caollaí, AWS User Group Ronan Guilfoyle, AWS Solutions Architect Lead 6:15 PM - LUCIANO MAMMINO, Senior Architect, fourTheorem "Serverless Podcast Transcription with Durable Lambda Functions, ECS Managed Instances, and Bedrock" 7:00 PM - NETWORKING 7:20 PM - GLEN KEANE, Senior Software Engineer, BRIGHTBEAM "Dr. Strangeagent, or: How I Learned to Stop Prompting and Love the State Machine” --Glen Keane, Senior Software Engineer, Brightbeam --Brian Finnegan, Principal Solutions Director, Brightbeam --Blake Rizk - Head of AI and Transformation, Laya Healthcare 8PM - Francis Flannery, Principal Solutions Architect, AWS Talk details TBA 8.30pm - ENDS SPEAKER BIOS +++++++++++++ fourTheorem LUCIANO MAMMINO Senior Architect at fourTheorem, AWS Serverless Hero, author, and podcast host, Luciano is a popular standout voice in the AWS Community, known for his immense expertise and tireless dedication to the community. Luciano's talk explores Lambda Durable Functions, ECS GPU workloads, Amazon Bedrock and building scalable, serverless podcast transcription pipelines. Luciano also co hosts AWS Bites with Eoin Shanaghy. The show has passed 150 episodes and, as a side project alongside a full time job, the production workflow has to be efficient. A few years ago Luciano and Eoin built an automated system to transcribe each episode and generate subtitles for YouTube and the website. That system became Podwhisperer, an open source pipeline that anyone can deploy in their own AWS account to automate podcast transcription. Three years is a long time in the AWS world. Better speech tooling showed up, new ways to run GPU workloads became practical, and durable orchestration in Lambda showed up as a great option to build workflows programmatically. So the project was rebuilt as Podwhisperer v2: a serverless transcription pipeline optimized for podcasts, with GPU accelerated transcription, speaker diarization, and an optional refinement step powered by Amazon Bedrock. v2 also produces captions in VTT, SRT, and JSON, including word level timing, so the output can be used directly across platforms without extra manual work. This talk is a practical case study of the v2 architecture and the trade offs behind it, but it is also designed to teach the foundations behind the building blocks that make it work. Luciano will introduce Lambda Durable Functions and explain how checkpointing and replay enable long running workflows to survive Lambda timeouts. He will then break down ECS Managed Instances and show how they can provision CUDA capable capacity on demand for containerized GPU workloads, scale to zero between runs, and avoid the operational overhead of managing an EC2 fleet. Finally, the talk will cover how Amazon Bedrock can be used for transcript refinement in a controlled way, improving readability without turning the output into creative writing. Along the way, the session digs into the implementation details that make or break these systems in practice, including idempotency, worker callbacks, queue based coordination, and startup latency when scaling from zero. Attendees will leave with a clear mental model of the services involved and a set of practical patterns they can apply to their own long running, event driven workloads. Speaker Bio Luciano is a Senior Architect at fourTheorem in Dublin, helping teams be successful with AWS by adopting serverless. He is an AWS Serverless Hero, co-host of the AWS Bites podcast, and co-author of the books Node.js Design Patterns and Crafting Lambda Functions in Rust. SPEAKER BIOS +++++++++++++ BRIGHTBEAM Blake Rizk - Head of AI and Transformation, Laya Healthcare: Blake is Head of AI and Transformation at Laya Healthcare, part of the AXA Group, where he leads the strategy and delivery of AI-driven transformation across the organisation. With over 20 years of experience in technology, data and digital transformation, he focuses on applying practical AI solutions to modernise core healthcare insurance operations and deliver measurable business value. Blake is currently responsible for advancing the adoption of artificial intelligence across many areas including claims processing, document intake, customer service, and knowledge management. Brian Finnegan - Principal Solutions Director, Brightbeam: Brian brings three decades of experience to his role as a Principal Solutions Director at Brightbeam. His expertise in technology delivery stems from a diverse background - which includes previous roles as a Technical Account Manager at AWS and Solutions Architect at Sungard and Presidio. As well as Cloud Architect at Zurich Insurance. This broad perspective gives Brian a pragmatic edge (cutting and otherwise) in bridging the gap between the promise of AI and other new technologies and the gritty reality of scaling it in production. Glen Keane \- Senior Software Engineer\, Brightbeam: Glen is a senior software engineer at Brightbeam who wears many hats. Lately, he’s been focused on building intelligent systems using AI. In his past he built smart systems without it. A fullstack developer in all the worst possible ways - he does backend, frontend and devops in both startup and enterprise environments. Previously, he’s been a node.js core maintainer at Nearform; the first engineering hire (outside the founding team) in an Irish Unicorn, Wayflyer; and a Head of Engineering for a local startup. He likes to pretend he’s cosplaying as Chewbacca when he needs a haircut (which is most of the time) and enjoys sitting by beaches in the sunny south east writing code and sipping coffee. BRIGHTBEAM - TALK DETAILS AI agents nail demos and drift in production. They skip steps, hallucinate results, and present confident garbage. The fix is not better prompts - it is better architecture. This talk shows how Step Functions patterns - typed I/O, explicit transitions, choice states, parallel execution, error catchers - solve the reliability problem for AI agents. I will walk through a real mutation testing workflow (Chaos Monkey for your test suite) that orchestrates five parallel AI agents, routes on data, and validates intent at the terminal state, without the orchestrating agent remembering anything. Three ideas that land: Split intelligence from execution. AI reasons freely to scope the problem, then a typed state machine drives the steps. Same mental model as Step Functions, applied to agents. Isolated sub-tasks. When a step needs real thinking, spin up a fresh agent with focused context and only the data it needs. The main thread stays deterministic. The sub-task gets the model's full attention. Self-driving I/O. Every step's output carries the next instruction. The agent follows the chain - it does not maintain a plan. If you have used Step Functions on AWS, you already understand the mental model. This talk shows where it maps directly to AI orchestration and where the AI-specific patterns diverge. Takeaways 1. Why agents drift and why prompt engineering cannot fix it 2. A concrete architecture you can implement the following week 3. Some core implementation patterns that have been tested in anger ++++++++++++++++++ SPEAK or SPONSOR We have speaking/sponsorship slots available! If you’re building something interesting on AWS we’d really love to hear from you. ++++++++++++++++++ SAVE THE DATE! - 24 March 2026! REINVENT RECAP + More!!! Don’t miss the next AWS community event featuring Major Announcements made @ REINVENT + More!!! Leading voices from the front lines of AWS / AI development AWS Ask an Architect All your questions about AWS answered at a single event! ———— Tickets are FREE but are Limited, so please RSVP Now. WANT TO SUPPORT? AWS User Group events offer a great opportunity to align your brand with AWS thought leadership and to reach a highly influential audience of senior engineers, architects, founders and technical decision-makers that traditional channels cannot match. The Dublin AWS User Group is amongst the longest established in the world and has a long reputation for delivering quality content and engagement opportunities. Meetups are run in a city centre live entertainment setting with full A/V capabilities including dedicated lighting and sound engineers and seating for 200. Contact us through the Messages Icon about Speaking and/or Supporting. Or use this form to find out about SPONSORING Agenda, may change slightly AWS User Group Dublin Team