Bill Genovese CISSP ITIL’s Post

This is a three-part post: When I read AI news, the sentiment often swings between fear and amazement - couple links below. I’m amazed by the technology, cautious about unintended consequences, and focused on building a product that enhances human work rather than undermining critical thinking as a cofounder of an AI-powered analytics tool for VoC data: valuearchitect.io. I see work changing in three big ways with augmented AI for the better: 1. Creating “Flow” in AI-Augmented Work: What if AI augmented our work so we, the knowledge workers achieved “flow”? As described by M. Csíkszentmihályi, flow is a state of deep enjoyment and creativity. Companies fostering this state have shown 5X productivity, check out HBR article below 👇 This is the type of application and experience creation that I am excited about. For example, in #userresearch, we love the flow of learning new things, but the synthesis of that research often feels overwhelming. AI can help by processing large amounts of data, allowing us to focus on insights and decision-making. This is the real value of augmented work—boosting critical thinking by reducing the cognitive load of data overload. This is an exciting time for #AIproductdevelopment, but it must be done with solid design principles: Care, Intention and Focus on enhancing human abilities. I'm curious about what my fellow creative leaders in workplace think: Picking on a few here: Laura Dye, Ozlem Brooke Erol, Julie Morgan, MBA, Michelle Batt, Megan Neese, Kirsty Nunez More on alignment in the next post. https://lnkd.in/g5g7RDqp https://lnkd.in/gVy2pZM9 https://lnkd.in/gsp-mqrZ https://lnkd.in/g642NSEU

Learn how #civilengineering can be enhanced with the power of #AI in today's #casestudy 👇 🔹Client Profile The US-based company on a mission to address architectural and engineering design challenges with the power of #GenAI. 🔹Challenges to address - design compliance issues - costly changes, derailing project timelines - access to design requirements and insights 🔹 Outcomes - enhanced access and user experience with our custom GenAI-powered platform for compliance check - better accuracy with integrated ASCE 7-16 design standards - incorporation of current IBC codes, tables, and figures in responses to user queries - hundreds of projects completed without timeline or budget overruns, as a result AI can revolutionize any industry - don't miss out on this chance to transform your business! Drop us a line 💬 -- #aiintegration #engineering #architecture #compliance #designcode #datascienceua

Late Chunking:- A new method for long-context retrieval in large-scale Retrieval-Augmented Generation (RAG) applications. Traditional methods either embed each document chunk independently or use advanced techniques like ColBERT with high storage and cost. Late chunking improves upon these by embedding the entire document first and then chunking the embeddings, preserving contextual information across chunks. Please follow these two articles for more details. https://lnkd.in/eXcgBzuj https://lnkd.in/dc68YzFs

Very useful industry use cases for #RAG and relevant types & architectures

Check out this blog from my colleague Dave Stewart Using your business or market terminology to explain EA and Business architecture concepts to stakeholders. And of course keeping things simple #enterprisearchitecture #modelling #MooD

In 2024 we will see more specialized LLMs which will lead to more accuracy and unlock many business use cases...

While LLMs often frustrate users by delivering generalized, outdated answers that erode user confidence in the technology, RAG-based models effectively address these shortcomings by providing timely, accurate, and contextually aware responses, enabling users to derive meaningful insights. But what exactly is RAG? How does it enhance the capabilities of LLMs in handling and transforming new data? Learn from our subject expert as he addresses these questions and explains how RAG can revolutionize enterprise workflows, offering a glimpse into its future potential. https://bit.ly/46eBEcv

Did you know? IDP can help you save up to 90% of document processing costs. Don't settle for it. Watch these IDP trends that will help you save more and boost ROI in 2025: https://lnkd.in/gXnkx4WX #IDPTrends #IDPTrends2025 #IntelligentDocumentProcessing #Trends2025

Thinking about the great post by Damien Benveniste, PhD about Agentic frameworks, and reflecting on discussions with multiple actors in the GenAI industry, 𝐈 𝐫𝐞𝐚𝐥𝐢𝐳𝐞𝐝 𝐡𝐨𝐰 𝐥𝐚𝐭𝐞𝐧𝐜𝐲, 𝐬𝐢𝐦𝐩𝐥𝐢𝐜𝐢𝐭𝐲, 𝐚𝐧𝐝 𝐠𝐨𝐯𝐞𝐫𝐧𝐚𝐧𝐜𝐞 𝐚𝐫𝐞 𝐜𝐫𝐢𝐭𝐢𝐜𝐚𝐥 𝐢𝐧 𝐢𝐧𝐝𝐮𝐬𝐭𝐫𝐢𝐚𝐥 𝐞𝐧𝐯𝐢𝐫𝐨𝐧𝐦𝐞𝐧𝐭𝐬. Customers are increasingly expecting simplicity, quick responses, and effective solutions tailored to their specific problems. While complex agentic frameworks are exciting and excel in research, planning, and idea generation, their over-engineering and internal dynamic complexity often make them less suitable for production environments where clarity and efficiency are essential. 𝐒𝐢𝐦𝐩𝐥𝐢𝐜𝐢𝐭𝐲 𝐧𝐨𝐭 𝐨𝐧𝐥𝐲 𝐞𝐧𝐚𝐛𝐥𝐞𝐬 𝐛𝐞𝐭𝐭𝐞𝐫 𝐠𝐨𝐯𝐞𝐫𝐧𝐚𝐧𝐜𝐞 𝐚𝐧𝐝 𝐬𝐲𝐬𝐭𝐞𝐦𝐚𝐭𝐢𝐜 𝐚𝐠𝐞𝐧𝐭 𝐝𝐞𝐩𝐥𝐨𝐲𝐦𝐞𝐧𝐭 𝐛𝐮𝐭 𝐚𝐥𝐬𝐨 𝐩𝐫𝐨𝐯𝐢𝐝𝐞𝐬 𝐚 𝐬𝐭𝐫𝐨𝐧𝐠 𝐟𝐨𝐮𝐧𝐝𝐚𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐜𝐫𝐞𝐚𝐭𝐢𝐧𝐠 𝐬𝐭𝐫𝐚𝐢𝐠𝐡𝐭𝐟𝐨𝐫𝐰𝐚𝐫𝐝 𝐦𝐮𝐥𝐭𝐢-𝐚𝐠𝐞𝐧𝐭 𝐬𝐭𝐫𝐚𝐭𝐞𝐠𝐢𝐞𝐬. These simpler approaches can address a wide range of business use cases while ensuring safe, reliable deployment. Tools like LangGraph or LlamaIndex Workflows, with their lightweight, event-based orchestration, are ideal for high-demand, latency-sensitive scenarios. As agentic systems evolve, they hold great promise for broader, more effective applications in the future. 🔗Damien post https://lnkd.in/eDRrM2Rs 🔗Industrial integration https://lnkd.in/e9CyuSSw / 🔗 From single to agentic https://lnkd.in/eJamWEra Image credit: https://lnkd.in/epAemShP 💬 What do you think?

“When engaging with any enterprise in the context of architecture-driven insight modelling, it’s important to learn and use the business dialect. Learning to speak their language develops an understanding of how they think and what they want from architecture, and how to apply the modelling discipline to deliver it.” Explore creative problem-solving and architecture modelling with our blog from our Senior MooD Technical Consultant Dave Stewart. https://lnkd.in/eKH_PFPN #MooD #ArchitectureModelling #Software #EnterpriseArchitecture