Your information to generative AI and ML at AWS re:Invent 2023
Sure, the AWS re:Invent season is upon us and as all the time, the place to be is Las Vegas! You marked your calendars, you booked your resort, and also you even bought the airfare. Now all you want is a few steering on generative AI and machine studying (ML) periods to attend at this twelfth version of re:Invent. And though generative AI has appeared in earlier occasions, this yr we’re taking it to the following stage. Along with a number of thrilling bulletins throughout keynotes, many of the periods in our observe will function generative AI in a single type or one other, so we will really name our observe “Generative AI and ML.” On this submit, we offer you a way of how the observe is organized and spotlight a number of periods we expect you’ll like. And though our observe focuses on generative AI, many different tracks have associated periods. Use the “Generative AI” tag as you’re looking the session catalog to seek out them.
The technical periods in our observe are divided into 5 areas. First, we’ll have a number of foundational periods associated to numerous elements of Amazon Bedrock—a totally managed generative AI service we launched earlier this yr. These will show you how to perceive the constructing blocks of your generative AI functions. Second, we’ll have periods masking the widespread generative AI use instances and functions. Right here you’ll even have an opportunity to find novel use instances and methods. Third, a lot of periods will likely be of curiosity to ML practitioners who construct, deploy, and operationalize each conventional and generative AI fashions. This yr, study LLMOps, not simply MLOps! Then, as we began doing final re:Invent, we’ll offer a number of periods on learn how to construct AI responsibly. The better the ability of newest transformer-based fashions, the better the accountability of all ML practitioners to do that proper. Be sure you take a look at the session on the simply launched PartyRock, an academic instrument for offering any builder with low-friction entry to study by means of experimentation in a basis mannequin playground constructed on Amazon Bedrock. And final however not least (and all the time enjoyable!) are the periods devoted to AWS DeepRacer!
Generative AI is on the coronary heart of the AWS Village this yr. Work together with a number of demos that function new functions, together with a contest that includes utilizing generative AI tech to pilot a drone round an impediment course. Speak with AWS consultants in 14 completely different industries and discover industry-specific generative AI use instances, together with demos from promoting and advertising, aerospace and satellite tv for pc, manufacturing, and extra. The Rising Tech Zone throughout the Expo options progressive startups that have been chosen into the AWS Generative AI Accelerator and the NVIDIA Inception 100 applications.
Should you’re new to re:Invent, you’ll be able to attend periods of the next varieties:
- Keynotes – Take part individual or just about and study all of the thrilling bulletins.
- Innovation Talks – Be taught in regards to the newest cloud expertise from AWS expertise leaders and uncover how these developments may also help you push your online business ahead. These periods will likely be livestreamed, recorded, and printed to YouTube.
- Breakout periods – These 60-minute periods are anticipated to have broad attraction, are delivered to bigger audiences, and will likely be recorded. Should you miss them, you’ll be able to watch them on demand after re:Invent.
- Chalk talks – Take pleasure in 60 minutes of content material delivered to smaller audiences with an interactive whiteboarding session. Chalk talks are the place discussions occur, and these give you the best alternative to ask questions or share your opinion.
- Workshops – In these hands-on studying alternatives, in the midst of 2 hours, you’ll have the ability to construct an answer to an issue, and perceive the inside workings of the ensuing infrastructure and cross-service interplay. Carry your laptop computer and be able to study!
- Builders’ periods – These extremely interactive 60-minute mini-workshops are carried out in small teams of fewer than 10 attendees. A few of these attraction to learners, and others are on specialised matters.
- NEW! Code talks – On this new session kind for re:Invent 2023, code talks are just like our widespread chalk speak format, however as an alternative of specializing in an structure answer with whiteboarding, the audio system lead an interactive dialogue that includes dwell coding or code samples. These 60-minute periods concentrate on the precise code that goes into constructing an answer. Attendees are inspired to ask questions and observe alongside.
When you’ve got reserved your seat at any of the periods, nice! If not, we all the time put aside some spots for walk-ins, so make a plan and are available to the session early.
That will help you plan your agenda for this yr’s re:Invent, listed here are some highlights of the generative AI and ML observe. So buckle up, and begin registering on your favourite periods.
Go to the session catalog to study all our generative AI and ML periods.
Keynotes
Adam Selipsky, Chief Government Officer, Amazon Internet Companies – Keynote
Tuesday November 28 | 8:30 AM – 10:30 AM (PST) | The Venetian
Be part of Adam Selipsky, CEO of Amazon Internet Companies, as he shares his perspective on cloud transformation. He highlights improvements in information, infrastructure, and synthetic intelligence and machine studying which might be serving to AWS clients obtain their targets sooner, mine untapped potential, and create a greater future.
Swami Sivasubramanian, Vice President of AWS Information and Machine Studying – Keynote
Wednesday November 29 | 8:30 AM – 10:30 AM (PST) | The Venetian
A robust relationship between people, information, and AI is unfolding proper earlier than us. Generative AI is augmenting our productiveness and creativity in new methods, whereas additionally being fueled by large quantities of enterprise information and human intelligence. Be part of Swami Sivasubramanian, Vice President of Information and AI at AWS, to find how you should use your organization information to construct differentiated generative AI functions and speed up productiveness for workers throughout your group. Additionally hear from buyer audio system with real-world examples of how they’ve used their information to help their generative AI use instances and create new experiences for his or her clients.
Innovation Talks
Dr. Bratin Saha, VP of AWS AI and ML Companies | AIM245-INT | Innovate sooner with generative AI
Wednesday November 29 | 1:00 PM – 2:00 PM (PST) | Venetian | Degree 5 | Palazzo Ballroom B
With the emergence of generative AI, we’re at a tipping level within the widespread adoption of machine studying. Be part of Dr. Bratin Saha, VP of AWS AI and ML Companies, to listen to how clients throughout industries are remodeling their enterprise with the newest breakthroughs in AI and ML, together with generative AI. Uncover the newest AWS improvements, hear from prime clients, and discover the place AI/ML is headed.
Francessca Vasquez, Vice President of Skilled Companies | ARC217-INT | From hype to impression: Constructing a generative AI structure
Wednesday November 29 | 11:30 AM – 12:30 PM (PST) | Venetian | Degree 5 | Palazzo Ballroom B
Generative AI represents a paradigm shift for the way corporations function at present. Generative AI is empowering builders to reimagine buyer experiences and functions whereas remodeling just about each {industry}. Organizations are quickly innovating to create the fitting structure for scaling generative AI securely, economically, and responsibly to ship enterprise worth. On this speak, learn the way leaders are modernizing their information basis, choosing industry-leading basis fashions, and deploying purpose-built accelerators to unlock the chances of generative AI.
Shaown Nandi, AWS Director of Expertise for Industries and Strategic Accounts | AIM248-INT | Unlocking the {industry} potential of generative AI
Wednesday November 29 | 4:00 PM – 5:00 PM (PST) | Venetian | Degree 5 | Palazzo Ballroom B
Generative AI has captured the creativeness of many industries and is poised to herald the following wave of technological developments. On this innovation speak, hear how the biggest industries, from healthcare and monetary providers to automotive and media and leisure, are utilizing generative AI to drive outcomes for his or her clients. Be part of Shaown Nandi, AWS Director of Expertise for Industries and Strategic Accounts, and {industry} leaders to listen to how generative AI is accelerating content material creation and serving to organizations reimagine buyer experiences.
Mai-Lan Tomsen Bukovec, Vice President, Expertise | AIM250-INT | Placing your information to work with generative AI
Thursday November 30 | 12:30 PM – 1:30 PM (PST) | Venetian | Degree 5 | Palazzo Ballroom B
How are you going to flip your information lake right into a enterprise benefit with generative AI? On this speak, discover methods for placing your proprietary datasets to work when constructing distinctive, differentiated generative AI options. Learn to make the most of your datasets utilizing Amazon SageMaker and Amazon Bedrock in addition to widespread frameworks like PyTorch with AWS compute, storage, and analytics. Hear finest practices for utilizing unstructured (video, picture, PDF), semi-structured (Parquet), and table-formatted (Iceberg) information for coaching, fine-tuning, checkpointing, and immediate engineering. Additionally hear completely different architectural patterns that clients use at present to harness their enterprise information for personalized generative AI options.
Breakout periods
AIM218 (LVL 200) | Construct your first generative AI utility with Amazon Bedrock
Monday November 27 | 2:30 PM – 3:30 PM (PST)
We’re really at an thrilling inflection level within the widespread adoption of ML with the expansion of generative AI functions. On this session, discover ways to construct your first generative AI utility with key providers equivalent to Amazon Bedrock. Get hints and suggestions for getting began quick, and see instance reference architectures for widespread use instances constructed with AWS AI and ML equivalent to self-service buyer help, textual content evaluation, report technology, post-call evaluation, and forecasting developments.
AIM225 (LVL 200) | Drive customized CX utilizing generative AI and Amazon Personalize
Tuesday November 28 | 5:00 PM – 6:00 PM (PST)
Delivering one of the best expertise is essential to seize and retain clients at present. With generative AI, it’s attainable to hyper-personalize focused suggestions for purchasing and streaming. Whereas normal taglines like “Individuals who purchased this additionally purchased . . .” or “Since you watched . . .” entice some, they don’t absolutely tackle particular person pursuits. Corporations should discover methods to dynamically generate compelling, extremely personalized content material. Amazon Personalize delivers capabilities powered by ML and generative AI to assist manufacturers create significant experiences. Be part of this session to listen to from powerhouse AWS media buyer FOX and learn the way hyper-personalized experiences can be utilized to construct engagement and drive income.
AIM327 (LVL 300) | Scaling FM inference to a whole bunch of fashions with Amazon SageMaker
Wednesday November 29 | 4:30 PM – 5:30 PM (PST)
Corporations want strong and cost-effective options to deploy basis fashions at scale. Moreover, SaaS suppliers want scalable and cost-effective methods to serve a whole bunch of fashions to their clients. This session explores learn how to use Amazon SageMaker to roll out a whole bunch of FMs cost-effectively at scale. Get an in depth overview of deployment methods to help large-scale generative AI inferencing for SaaS, and discover ways to architect options that maximize scaling capabilities for efficiency and price.
AIM333 (LVL 300) | Discover text-generation FMs for prime use instances with Amazon Bedrock
Tuesday November 28| 2:00 PM – 3:00 PM (PST)
Basis fashions can be utilized for pure language processing duties equivalent to summarization, textual content technology, classification, open-ended Q&A, and data extraction. With Amazon Bedrock, you’ll be able to select highly effective FMs from AI21 Labs, Anthropic, and Cohere to seek out the fitting FM on your use case such because the Jurassic-2, Claude, and Command households of text-generation FMs. Be part of this session to study which FM is finest suited on your use case.
AIM332 (LVL 300) | Discover picture technology and search with FMs on Amazon Bedrock
Thursday November 30 | 11:00 AM – 12:00 PM (PST)
Basis fashions perceive a number of types of enter, equivalent to photographs and texts. Be part of this session to discover ways to construct transformational experiences utilizing photographs in Amazon Bedrock.
AIM377 (LVL 300) | Immediate engineering finest practices for LLMs on Amazon Bedrock
Monday November 27 | 9:00 AM – 10:00 AM (PST)
Immediate engineering is the method of guiding giant language fashions to provide desired outputs. On this session, get an summary of immediate engineering finest practices and discover ways to select essentially the most applicable codecs, phrases, phrases, and symbols to get essentially the most out of generative AI options whereas enhancing accuracy and efficiency. This session makes use of the Claude 2 LLM for instance of how immediate engineering helps to unravel advanced buyer use instances. Additionally learn the way prompts could be built-in together with your structure and learn how to use API parameters for tuning the mannequin parameters utilizing Amazon Bedrock.
Chalk talks
AIM341 (LVL 300) | Ship personalized search capabilities utilizing Amazon Bedrock
Wednesday November 29 | 5:30 PM – 6:30 PM (PST)
Vector embeddings are numerical representations of your textual content, picture, audio, and video information that can be utilized to know the connection between sentences or phrases to seek out extra related and contextual data in response to a consumer question. Embeddings could be saved in a database and are used to allow streamlined and extra correct searches. You should utilize an embeddings mannequin in Amazon Bedrock to create vectors of your group’s information, which might then be used to allow semantic search. Be part of this hands-on chalk speak to learn the way.
AIM340-R (LVL 300) | Customise your FMs securely to ship differentiated experiences
Wednesday November 29 | 6:00 PM – 7:00 PM (PST)
Basis mannequin customizations show you how to construct differentiated generative AI functions utilizing your personal information. It’s simple to securely customise fashions in Amazon Bedrock. You possibly can level Amazon Bedrock at a number of labeled examples in Amazon S3, and the service can fine-tune the FM for a selected job with out having to annotate giant volumes of knowledge; none of your information is used to coach the unique base FMs. Be part of this chalk speak for a deep dive on FM customizations by means of an interactive demo.
This session will likely be repeated Thursday, November 30 11:00 AM – 12:00 PM (PST), and Friday, December 1 8:30 AM – 9:30 AM PST.
AIM342 (LVL 300) | Advancing accountable AI: Assessing and mitigating threat
Wednesday November 29 | 4:30 PM – 5:30 PM (PST)
Threat evaluation is a vital a part of growing AI options responsibly, particularly with rising {industry} requirements and legal guidelines relating to AI threat, equivalent to ISO 42001 and the EU AI Act. This chalk speak offers an introduction to finest practices for threat evaluation associated to equity, robustness, explainability, privateness and safety, transparency, and governance. Discover examples to estimate the severity and probability of potential occasions that could possibly be dangerous. Study Amazon SageMaker tooling for mannequin governance, bias, explainability, and monitoring, and about transparency within the type of service playing cards as potential threat mitigation methods.
AIM347-R (LVL 300) | Subsequent-generation ML builder expertise
Thursday November 30 | 4:00 PM – 5:00 PM (PST)
Amazon SageMaker affords completely different built-in improvement environments (IDEs) which might be purpose-built for machine studying. On this chalk speak, discover ways to choose and use your most popular atmosphere to carry out end-to-end ML improvement steps, from getting ready information to constructing, coaching, and deploying your ML fashions. Uncover how one can shortly add information, create new notebooks, prepare and tune fashions, transfer backwards and forwards between steps to regulate experiments, collaborate seamlessly inside your group, and deploy fashions to manufacturing multi function place.
This session will likely be repeated Friday, December 1 9:00 AM – 10:00 AM (PST), and Friday, December 1 11:30 AM – 12:00 PM (PST).
AIM352-R (LVL 300) | Securely construct generative AI apps and management information with Amazon Bedrock
Monday November 27 | 11:30 AM – 12:30 PM (PST)
Generative AI functions have captured widespread consideration; nevertheless, they’ve additionally launched new safety challenges, particularly across the dealing with of buyer information. Organizations need to be certain that their information stays secure and safe whereas working with basis fashions and don’t need to fear about their information getting used to coach an FM. Amazon Bedrock offers complete information safety and privateness. On this chalk speak, discover architectures, information flows, and security-related elements of mannequin fine-tuning, in addition to prompting and inference, when you study Amazon Bedrock’s safety capabilities.
This session will likely be repeated Wednesday, November 29 6:00 PM – 7:00 PM (PST), and Thursday, November 30 4:00 PM – 5:00 PM (PST).
AIM404 (LVL 400) | Prepare and deploy FMs on Amazon EC2 and Amazon SageMaker, feat. Flip AI
Wednesday November 29 | 2:30 PM – 3:30 PM (PST)
Organizations which might be operating machine studying methods and generative AI functions on their native laptops/servers need to benefit from the scalability and efficiency of the AWS Cloud. On this chalk speak, hear about compute and ML providers from self-managed Amazon EC2 to completely managed Amazon SageMaker that you should use to construct, prepare, and deploy basis fashions. See a demo of how one can fine-tune a Steady Diffusion mannequin on Amazon EC2 after which deploy it on SageMaker utilizing the AWS Deep Studying AMIs (DLAMI) and AWS Deep Studying Containers. Additionally, hear how Flip AI constructed their very own fashions utilizing these AWS providers.
Workshops
AIM302 (LVL 300) | Use generative AI to extract insights from contact heart recordings
Monday November 27 | 8:30 AM – 10:30 AM (PST)
Learn to derive insights from contact heart recordings and different media utilizing Amazon Transcribe and generative AI. On this workshop, discover ways to mix automated name recording, transcription, post-call evaluation, sentiment evaluation, problem detection, and name summarization from your personal telephony recordings (Cisco, Genesys, Talkdesk, Avaya, and extra) utilizing AWS Contact Middle Intelligence (CCI) options and generative AI. See demos on learn how to construct analytics dashboards and integrations between LLMs and Amazon QuickSight to visualise your key metrics. You will need to deliver your laptop computer to take part.
AIM307 (LVL 300) | Retrieval Augmented Era with Amazon Bedrock
Wednesday November 29 | 8:30 AM – 10:30 AM (PST)
Massive language fashions are sometimes restricted by the information they have been educated on and don’t all the time present up-to-date responses—or worse, they make issues up. To beat this limitation, you’ll be able to complement prompts with up-to-date data utilizing embeddings saved in vector databases, a course of often called Retrieval Augmented Era (RAG). With supplemental data within the immediate offering extra context, the LLM can reply extra precisely and is much less more likely to hallucinate. On this workshop, discover ways to use vector databases with Amazon Bedrock, a service that makes basis fashions from Amazon and main AI corporations out there through a single API. You will need to deliver your laptop computer to take part.
AIM304 (LVL 300) | How you can generate textual content responsibly utilizing basis fashions on AWS
Wednesday November 29 | 5:30 PM – 7:30 PM (PST)
Basis fashions equivalent to Claude are generally used to create new items of authentic content material, equivalent to brief tales, essays, social media posts, and webpage copy, and likewise to summarize textual content from articles, weblog posts, books, and extra. On this workshop, study how one can generate textual content in minutes utilizing basis fashions out there by means of Amazon Bedrock in a accountable means. You will need to deliver your laptop computer to take part.
Code talks
AIM364-R (LVL 300) | Enhance ML improvement with Amazon SageMaker Studio notebooks
Tuesday November 28 | 4:00 PM – 5:00 PM (PST)
Amazon SageMaker Studio notebooks are collaborative notebooks which you could launch shortly and that may show you how to combine with purpose-built ML instruments in SageMaker and different AWS providers for full ML improvement. On this code speak, discover ways to put together information at scale utilizing built-in information preparation help, co-edit the identical pocket book in actual time, and automate conversion of pocket book code to production-ready jobs. This speak additionally introduces the brand new generative AI-powered options that may show you how to maximize productiveness, write higher-quality code, and enhance safety.
This session will likely be repeated Wednesday, November 29 12:00 PM – 1:00 PM (PST).
Builders’ periods
AIM219-R (LVL 200) | Be taught and experiment with LLMs in Amazon SageMaker Studio Lab
Monday November 27 | 10:00 AM – 11:00 AM (PST)
Machine studying can sound sophisticated, however Amazon SageMaker Studio Lab makes it simpler for anybody to get began for free of charge. On this hands-on builders’ session, be guided by means of the fundamentals of experimenting with giant language fashions in Amazon SageMaker Studio Lab. No prior machine studying expertise is required. You will need to deliver your laptop computer to take part.
This session will likely be repeated Monday, November 27 4:00 PM – 5:00 PM (PST), Tuesday, November 28 3:30 PM – 4:30 PM (PST), Wednesday, November 29, 12:00 PM – 1:00 PM (PST), and Thursday, November 30 11:30 AM – 12:30 PM (PST).
AWS DeepRacer
Get able to race with AWS DeepRacer at re:Invent 2023!
Builders, fasten your seatbelts—AWS DeepRacer is bringing ML to everybody at re:Invent! Whether or not you’re trying to get began with ML or enhance your expertise, AWS DeepRacer affords an thrilling technique to get hands-on with ML.
Watch the world’s prime 72 racers of the AWS DeepRacer 2023 League battle it out Monday by means of Wednesday at our Championship Stadium on the Venetian Expo. It can all come right down to the finale on Wednesday (November 29) at 2:30 PM (PST) because the eight finalists compete for the cup and $44,000 in prize cash. You may also get behind the wheel your self on November 30, when the observe opens for the 2024 Open Racing. Publish the quickest time and also you’ll win a ticket again to Vegas for the 2024 Championship!
Dive into 10 not-to-miss workshops the place you’ll study to coach reinforcement studying fashions, clear up enterprise issues with generative AI, and extra. Need to study suggestions and methods from one of the best racers on this planet? Be sure you take a look at our DPR301 workshop that includes 5 of our prime AWS DeepRacer League Champions who will likely be sharing their approaches for coaching their AWS DeepRacer fashions and answering questions throughout an open Q&A.
Don’t neglect to take a look at the remainder of the AWS DeepRacer workshops earlier than they fill as much as reserve your spot! Whether or not you’re taking a workshop, take a spin in our gamified digital racing simulator, catch the worldwide competitors, or take a look at your personal ML mannequin on the observe, AWS DeepRacer brings the joys of high-speed racing to hands-on machine studying at re:Invent. Let the countdown start. We are able to’t wait to see you in Las Vegas!
See you at re:Invent!
Ensure that to take a look at the re:Invent content catalog and the generative AI at re:Invent guide for extra gen AI and ML content material at re:Invent. We’ll see you there!
Concerning the authors
Denis V. Batalov is a 17-year Amazon veteran and a PhD in Machine Studying, Denis labored on such thrilling initiatives as Search Contained in the E-book, Amazon Cellular apps and Kindle Direct Publishing. Since 2013 he has helped AWS clients undertake AI/ML expertise as a Options Architect. At present, Denis is a Worldwide Tech Chief for AI/ML chargeable for the functioning of AWS ML Specialist Options Architects globally. Denis is a frequent public speaker, you’ll be able to observe him on Twitter @dbatalov.
Paxton Corridor is a Advertising and marketing Program Supervisor for the AWS AI/ML Group on the AI/ML Schooling workforce at AWS. He has labored in retail and experiential advertising for the previous 7 years, centered on growing communities and advertising campaigns. Out of the workplace, he’s captivated with public lands entry and conservation, and enjoys backcountry snowboarding, climbing, biking, and climbing all through Washington’s Cascade mountains.
