Self-paced Day 1 session focusing on AI foundations and the power of large language models.

Self-paced session focusing on AI foundations and large language models.

Vanilla Retrieval-Augmented Generation (RAG) is becoming more and more adopted - but What is next? Join us for an introduction to Generative Feedback Loops (GFL). GFLs can level-up your RAG architecture by generating more insights on top of your data. This empowers you to implement smarter chatbots, agents, and other AI-driven solutions. Based on hands-on examples, we will explore the following questions: a) What are GFLs are? b) How do GFLs work? c) What challenges and use-cases can we tackle with GFLs; and d) How can I define my AI workflows to implement GFLs. Learn about the next level of AI applications and join us for a Live Demo.

Overview of how Small Language Models (SLMs) can solve business problems and boost device intelligence, with a comparison to Large Language Models (LLMs). Focus on fine-tuning and customization, edge deployments for real-time processing, and a cloud-to-edge deployment roadmap using Azure to maintain IP ownership and integration. Includes perspectives on leveraging SLMs to revolutionize devices and mentions Microsoft’s Student Ambassador Program energizing students to advance intelligent tech.

Large language models unlock exciting new opportunities - but there's still lots of hard work to do for engineers, designers and product managers to turn foundation models into products users love! In this session, you'll learn about GitHub's journey building Copilot code review, currently in public preview, and the lessons learned along the way.

During the session, we’ll discuss the challenges that prompt engineering has presented, both when it first gained popularity and as it continued to evolve. We’ll share how these challenges informed the development of our prompt engineering tooling and workflows. We’ll cover: Standardizing communication with LLMs; Using templating to customize prompts; Building prompt-centric production workflows; Working with structured LLM output; Ensuring the quality of LLM output; Creating tooling that supports our prompt engineering workflows.

One of the latest innovations in Generative AI (GenAI) technology are Large Language Models (LLMs). You can exploit LLMs to solve many tasks that were incredibly challenging not that long ago, and some say you can solve anything with GenAI. But can you? Recently, we decided to reshape an old application for fraud detection by introducing AI in it. This talk is a summary of both our successful and failed attempts to solve everything with AI.

In this talk, I will share our journey, which began with local applications leveraging Open-Source LLMs and has now evolved to serve our internal teams. I will dive into the practical lessons we've learned along the way, offering insights that can help you navigate the complexities of LLM implementation.

Organizations develop feedback loops to continuously enhance quality. One such loop is the learning from user interactions with your data, retraining models, deploying new models and learning again. The learning curve to create a loop like this is steep, it requires ML experience and tools. However, most teams can easily provide labeled examples. In-Context Learning (ICL) is a method to add classification examples as input to foundation models (like LLMs).\nThis talk defines an Adaptive ICL strategy using Retrieval for Examples, where the output is used for content retrieval, example set expansion for future model training and real-time user behaviour tracking. Adaptive ICL is hence an easy way for teams to get immediate results with AI, while laying the foundation for more advanced ML loops in the future.

Fireside chat on evaluating quality in LLM models pre-production.

LLMs have unlocked new opportunities in NLP with their possible applications. Features that used to take months to be planned and developed now require a day to be prototyped. But how can we make sure that a successful prototype will turn into a high-quality feature useful for millions of customers? In this talk, we will explore real examples of the challenges that arise when ensuring the quality of LLM outputs and how we address them at Grammarly.

This project aims to leverage retrieval augmented generation (RAG) and fine-tuning of LLMs to create an AI-based assistant for mental health, which could be used to support a psychotherapist.

Fundamentals of Generative AI and LLMs; best practices for building AI solutions on AWS; overview of AWS services like SageMaker; real-world applications and case studies.

Discover the basics of generative AI, including core models, functionalities. Learn how to utilize these models to build your own generative AI applications.

In this talk, we will delve into the emerging challenges that generative artificial intelligence GenAI and Large Language Models (LLMs) bring to the world of software quality and testing. We will explore how the integration, use, or design of solutions with GenAI models pose new challenges, including data quality and bias detection, dealing with non-determinism in our test automation and privacy and security concerns. We will review some proposals that are being implemented to adjust our testing tasks to the development of such systems, including approaches to automated testing and observability.

An intro to RAGHack, a global hackathon to develop apps using LLMs and RAG. A large language model (LLM) like GPT-4 can be used for summarization, translation, entity extraction, and question-answering. Retrieval Augmented Generation (RAG) is an approach that sends context to the LLM so that it can provide grounded answers. RAG apps can be developed on Azure using a wide range of programming languages and retrievers (such as AI Search, Cosmos DB, PostgreSQL, and Azure SQL). Get an overview of RAG in this session before diving deep in our follow-up streams.

LLM Tech Talk + Gemini API Cookbook Hands-on Technical workshop

In this session, we will discuss how to identify high stakes data, and share different ways of refining high-stakes unstructured data into meaningful, actionable data, and leverage the AI, LLMs to detect churn, high quality leads, and map a user's message to a user's journey and fill in the blanks.

LLMs like GPT can give useful answers to many questions, but there are also well-known issues with their output: The responses may be outdated, inaccurate, or outright hallucinations, and it’s hard to know when you can trust them. And they don’t know anything about you or your organization private data (we hope). RAG can help reduce the problems with “hallucinated” answers, and make the responses more up-to-date, accurate, and personalized - by injecting related knowledge, including non-public data. In this talk, we’ll go through what RAG means, demo some ways you can implement it - and warn of some traps you still have to watch out for.

Julian and David will cover the Hackathon project they worked on that won at the New York Stock Exchange— fine tuning an LLM to generate summaries for airflow task failures.