AI Con USA 2025 - AI Engineer
Customize your AI Con USA 2025 experience with sessions covering AI engineering.
Sunday, June 8
Fundamentals of AI—ICAgile Certification (ICP-FAI)
Monday, June 9
Beginning Data Analysis and Machine Learning with Jupyter Notebooks
In this beginner-friendly workshop you'll see how you can get started with data analytics and data science using Jupyter Notebooks. Matt will start with the basics of notebooks and then move on to using Python, Pandas, and NumPy to perform basic exploratory data analysis. See how you can use Plotly Express to create interactive charts and visuals with only a minimal amount of code. Once you've grasped the basics of understanding and visualizing the data Matt will move on to machine learning with SciKit-Learn as you train and evaluate predictive regression and classification models. The...
Get Your Data Ready for AI/ML
Understanding the readiness of your source data before you launch an expensive AI/ML project lets you take corrective data engineering measures that will streamline the project and give you the best probability of a successful outcome. Artificial Intelligence (AI) and Machine Learning (ML) projects can provide significant returns on investment when they are applied to narrow but difficult business problems and are supported by adequate amounts of relevant, quality data. Many such projects start with high hopes but get derailed due to fundamental problems with source data, which were...
Tuesday, June 10
Clean Your Filthy RAGS! Optimizing, Accelerating, and Evaluating RAG Applications
NewRetrieval-Augmented Generation (RAG) applications are becoming essential for companies, combining AI with real-time data retrieval to enhance customer experiences. While Large Language Models (LLMs) handle general conversation well, they struggle with domain-specific, up-to-date information, often producing inaccurate or unhelpful responses. This workshop will empower participants with the necessary skills to optimize RAG applications using existing best practices. Justin will walk through integrating RAGAS, a framework designed to evaluate, monitor, and fine-tune the performance of RAG...
Common Software, Data, and AI Architectures and the Ways They Fail
NewThis tutorial will examine the complexity that lives in software systems, data ingestion workflows, MLOps pipelines, and artificial intelligence systems. This session blends together cloud architecture, quality assurance, risk management, and security mindsets as Matt Eland explores how modern systems are structured, the problems their complexity helps us solve, and the ways these systems can break - or be broken. The session will alternate between interactive lectures with practical illustrations and group exercises around case studies as you explore how existing systems can fail and what...
Image Classification using LLMs and Transfer Learning
Large language models, like Generative Pre-trained Transformers (GPTs) to create textual content in ChatBots and other Generative AI applications, have garnered much attention recently. However, not all data is textual. Another important use of LLMs is for image processing to address real-world problems such as: real-time object recognition, classification of images, and generation of modern art. Jeffery Payne will explore how large language models for images (also known as visual language models - VLMs) can be used for image classification, object detection, image capture generation, and...
Responsible AI: Building Smarter, Ethically Empowered Tech
In an era of artificial intelligence (AI), many leaders have become concerned with the potential consequences of using AI in their products. Scott Peterson has met with many organizations about Artificial Intelligence. Their common question is, “ Is AI safe to use within our organization?” and “How do we begin implementing a plan around how we can benefit from it?” AI is not only shaping the future of technology but it’s also redefining the ethical landscape of innovation, the "RAI Revolution" session at AICon stands as a beacon for professionals seeking to harmonize technical prowess with...
Wednesday, June 11
Real-World Open Source Agentic Testing Framework: TestZeus
Test automation is evolving from manual scripting to AI-driven collaboration, where AI agents work together to ensure reliable test execution. This presentation introduces TestZeus, an open-source framework that leverages LLMs to transform Cucumber/Gherkin specifications into automated tests. Robin Gupta will explore its architecture and demonstrate how it handles common challenges like shadow DOMs and iframes. Learn how to implement and debug this next-generation framework, positioning yourself at the forefront of intelligent test automation. Join Robin to discover how TestZeus can...
Powering Complex Solutions with Agentic Systems
This session explores the revolutionary potential of agentic systems—autonomous agents equipped with diverse tools to tackle multifaceted tasks effectively. By integrating tools, agentic workflows enable intelligent, adaptive solutions to complex challenges. Apurva will start with an introduction to agentic systems, highlighting their ability to utilize various tools dynamically to achieve goals with minimal human intervention. Through real-world case studies, she will demonstrate how tools like APIs, databases, and external services are orchestrated within these systems to simulate...
Unlocking Agentic AI Superpowers in the Quantum Multiverse
Step into the future of AI with an immersive dive into the cutting-edge world of AI-Augmented GenAI and how it intercepts with Quantum technologies. This session explores the intersection of multi-threaded AI workflows and the powerful anomaly detection capabilities made possible by Quantum-inspired computing. Discover the latest trends in ModelOps, where agile deployment strategies for domain-specific generative AI models are accelerating innovation in industries ranging from healthcare to defense. Jonathon will uncover how GraphRAG (Graph Retrieval-Augmented Generation) is...
Moving from Reproducible Data Science to the World of LLM, GPT, RAG, and Hallucination
We are moving from the world of reproducible data science (machine learning, area under the curve, a/b testing, model selection, model testing, model fitting and model drift to find the best way to run a data science organization) to being faced with the "Easy Button." This "Easy Button" is often wrong, has poor data protocols, and tries to please; without factual basis. Join Bob to explore ways to make the world better in the bridge from science to the new world of utterances. This session will include examples and experiences from IBM Watson, AI at GE Healthcare, AI at Microsoft, and...
Customer Churn Prediction Using MLFlow and Streamlit
This session will guide you through every stage of the machine learning process, from data preprocessing and feature engineering to model training, pipeline construction, and deployment. The session will begin by preparing and transforming data to ensure it’s ready for model training. Next, dive into building scalable ML pipelines within MLFlow, where you’ll learn how to track experiments, monitor model performance, and manage version control. These features enable a streamlined and reproducible workflow, empowering both beginner and experienced practitioners to understand and implement...
Decoding the Black Box: Unlocking LLM Observability
LLMs are revolutionizing AI, but their complexity and black-box nature can make it challenging to understand their behavior and ensure optimal performance. This session will demystify LLM observability, providing practical insights and best practices for gaining visibility into LLM interactions, performance, and security. Learn how to: discover the essential metrics to monitor for LLM health and performance, implement effective logging and tracing strategies to track LLM requests and identify bottlenecks, optimize LLM inference performance using profiling techniques, proactively detect and...
Mastering Prompt Engineering: Unlocking the Power of Prompt Patterns and Meta-Prompting Technique
As AI becomes increasingly central to professional workflows, the ability to craft effective prompts has emerged as a crucial skill. While many users struggle with trial-and-error approaches, this session introduces a systematic framework for prompt engineering based on proven patterns and and meta-prompting techniques. You'll learn how to apply specific prompt patterns such as persona, meta-language creation, and flipped interaction, and more, to achieve consistently better results. You will then explore meta-prompting - a powerful technique where AI assists in generating, refining, and...
Thursday, June 12
RAG at Scale: Building Production-Ready GenAI Solutions
This session serves as a deep dive into the strategies and best practices for data scientists aiming to build, fine-tune, and scale Retrieval Augmented Generation (RAG) based Generative AI (GenAI) applications. The core objectives for data scientists in the AI development cycle center around providing highly relevant results for end users and maintaining cost-effectiveness to support the sustainable growth of their products. RAG is a pivotal method that enables GenAI applications to operate effectively on proprietary data. At the heart of RAG are robust retrieval systems, which are crucial...
Continuous Testing for AI Applications
In the era of artificial intelligence, software testing has evolved from a finite phase in development to an ongoing, dynamic process of monitoring. Unlike traditional deterministic systems, AI-driven applications operate probabilistically, introducing variability and uncertainty in outputs even with consistent inputs. This paradigm shift requires a rethinking of testing strategies, moving towards continuous monitoring to ensure performance, fairness, and reliability in production environments. This session will explore how QA teams can integrate AI-specific methods, anomaly detection, and...
RAG Has Evolved - Enhance Your RAG Pipeline with These Concepts
The majority of businesses today can set up a fundamental RAG pipeline that effectively handles most use cases. However, this basic setup eventually reaches its limitations in terms of functionality and accuracy, hindering further advancements. Matt Payne aims to detail the necessary pipeline components for building advanced RAG pipelines. For each component, he will explain the what, when, why, and how and provide real-world examples. Key areas of focus include leveraging tools and function calling, which enables you to create a systematic approach to using knowledge from multiple sources...
Navigating Compliance with the EU AI Act: Strategies for Responsible AI Development and Governance
As AI systems continue to transform industries, the EU AI Act sets a pivotal regulatory framework to ensure responsible and ethical use of artificial intelligence across Europe. This session will provide practical insights into navigating compliance with the EU AI Act, addressing key requirements for high-risk AI systems, transparency, accountability, and data governance. Joanna will explore how organizations can align their AI development processes with the Act's standards while maintaining innovation. Attendees will gain actionable strategies for integrating compliance into existing...