Deep Analysis

💡 The Shift to AI Agents: Prepare for Pay-Per-Use Pricing 💡 The free AI buffet is ending, and metered pricing is on the rise. Companies are moving beyond generative AI assistants like ChatGPT and Co-Pilot to AI agents that can make decisions and perform tasks autonomously. But here’s the catch: you’ll soon be paying for every interaction. Salesforce is leading the charge, with plans to charge $2 per conversation for its new AI agents. The shift toward consumption-based pricing is about to change the landscape for companies relying on AI. 🧠 Key Takeaways: AI agents are more powerful than generative assistants. They don’t just assist—they make decisions and follow workflows. ⚡Prepare for metered costs. Instead of flat fees, every AI-driven interaction could come with a price tag. ⚡Identify where AI agents bring real ROI. Not every use case will justify the cost, so choose wisely. ⚡Budgeting for AI will become more complex. Companies will need to plan for AI costs like they plan for cloud services. Is your organization ready to manage the new economics of AI? The shift from free trials to usage-based pricing is happening fast. This is a quick snippet from We Love Ugly Data Podcast, Series 3, Episode 9. Head over to our YouTube channel to subscribe. #AI #AIagents #Salesforce #genAI #DigitalEconomy #BusinessStrategy

Don’t Settle for the Usual Suspects in #IDP #Document Processing

Why AI Agents are the New Priority for Enterprises 👉 Agents vs. Assistants: AI agents can perform tasks autonomously, while assistants help with language or simple tasks. Agents' ability to complete actions makes them more valuable in enterprise settings. 👉 ROI Simplicity: Measuring ROI is easier with agents, as each task they complete has a clear, measurable cost. For example, a $2-per-task model is more straightforward than calculating vague productivity gains from assistants. 👉 Autonomy and Cost Efficiency: Agents can reduce the need for human intervention by automating end-to-end processes. This not only speeds up workflows but also lowers the potential for costly human errors. 👉 Salesforce's Hard Pivot: Salesforce is aggressively pushing this shift toward agents, signaling a broader trend. Their upcoming product line will likely feature AI agents heavily, charging customers based on agent-driven tasks. [Stay tuned for further commentary here, after the annual Salesforce event in September.] Takeaway: AI agents provide clearer ROI and better cost efficiency by autonomously completing tasks, making them a priority for businesses focused on automation and streamlined operations. This is a quick overview of a blog post from Matt Mullen, Here Are the Agents, They've Come to Collect. Follow us here or head over to our website to subscribe to our blog and never miss our unique insights. #AgenticAI #Salesforce #genAI #chatbots #businessautomation

As 2024 rounds the bend to its end, here's a look back at our predictions for this year. We think we hit more than missed (and we'll do a recap soon - and look for our 2025 predictions coming up too). How do you think we did? 1. Tech investors will focus more closely on profit and business value. 2. Generative AI will face its first cold winter. 3. We'll see an explosion of GenAI startups. 4. RPA will bounce back. 5. Cross-lingual conversations will start to take off. 6. Knowledge graphs and data meshes will gain traction. 7. Document management will become cool again. 8. For the first time in human history, machines will read and process more documents than knowledge workers do. 9. IDP software companies will replace large chunks of existing product platforms with new generative AI-powered platforms. 10. For citizen developers, GenAI will replace low-code/no-code platforms. 11. Predictive AI will make a comeback. What'd we get right? Wrong? For longer explanations, read the attached document. Looking ahead to 2025, what do you see coming?

The Future of Cloud Software Pricing is Metered 𝐒𝐡𝐢𝐟𝐭 𝐟𝐫𝐨𝐦 𝐀𝐬𝐬𝐢𝐬𝐭𝐚𝐧𝐭𝐬 𝐭𝐨 𝐀𝐠𝐞𝐧𝐭𝐬: AI agents are overtaking generative AI assistants as the primary focus for companies like Salesforce. Agents can handle more complex tasks autonomously, which justifies a different pricing model. 𝐌𝐞𝐭𝐞𝐫𝐞𝐝 𝐏𝐚𝐲𝐦𝐞𝐧𝐭 𝐌𝐨𝐝𝐞𝐥: We’re transitioning from flat-rate, all-you-can-use models to pay-per-use systems. This metered billing allows companies to pay based on how much they actually use AI tools. 𝐒𝐚𝐥𝐞𝐬𝐟𝐨𝐫𝐜𝐞 𝐋𝐞𝐚𝐝𝐢𝐧𝐠 𝐭𝐡𝐞 𝐂𝐡𝐚𝐫𝐠𝐞: Salesforce has already started using consumption-based pricing for some of its services, like Data Cloud. This move signals that other cloud software providers may soon follow suit. 𝐓𝐫𝐚𝐜𝐤𝐢𝐧𝐠 𝐔𝐬𝐚𝐠𝐞 𝐢𝐬 𝐄𝐬𝐬𝐞𝐧𝐭𝐢𝐚𝐥: With metered pricing, businesses will need to monitor and plan their usage carefully. The unpredictability of consumption could lead to unexpected costs if not managed properly. 𝐓𝐚𝐤𝐞𝐚𝐰𝐚𝐲: The future of AI and cloud software is in pay-per-use models, and companies need to be prepared to manage consumption closely. This is a quick overview of a blog post from Matt Mullen, Here Are the Agents, They've Come to Collect. Follow us here or head over to our website to subscribe to our blog and never miss our unique insights. #AgenticAI #genAI #Salesforce #chatbots

We Love Ugly Data! – The Deep Analysis Podcast: Series 3, Episode 6

AI Agents: Are We Ready for Autonomy? Dan Lucarina might have AI derangement syndrome. What do you think? The buzz around autonomous AI agents is growing. They promise to transform our work by handling mundane tasks without human input. But are we truly ready? A recent incident with Alexa made me question this trend. While driving, I asked Alexa to play a song. It misunderstood and subscribed me to a paid service without confirmation. This raised several concerns: Should AI have default permissions to make purchases? Is an "are you sure" step necessary before AI takes action? Who's responsible when AI decisions cause problems? The implications for business are significant. Trusting AI with autonomy requires a new level of faith in machines. We're not there yet. Current AI, like GenAI chatbots, still makes frequent errors. Vigilance is key to catch and correct these mistakes. Imagine the potential issues in critical business processes. Final thought: While we shouldn't fear progress, we must approach AI autonomy with caution. Default permissions and lack of confirmation steps could lead to unforeseen consequences. It's time to have serious discussions about AI boundaries and safeguards. Head over to our website to read Dan's thoughts in his blog post on this topic: Better to beg forgiveness later? No way, AI #genAI #AI #risk #chatbots

IDP is a very crowded field

Agentic AI: The next "Killer App" for Artificial Intelligence after GenAI? We’ve barely caught our breath from the explosion of Generative AI, and now there’s another disruptor on the horizon: Agentic AI (or "agent-based AI," if “agentic” doesn’t roll off the tongue). Unlike GenAI, which creates, Agentic AI acts—handling tasks end-to-end and making real-time decisions in ways we never thought possible. Here’s the gist: 1️⃣ Beyond RPA: Remember when Robotic Process Automation (RPA) was the hottest thing in automation? It revolutionized repetitive tasks. But RPA hits a wall when exceptions arise. Enter Agentic Process Automation (APA), powered by Agentic AI, designed to handle decision-heavy tasks without needing constant human intervention. 2️⃣ LAMs—The Brains Behind the Agents: Agentic AI uses Large Action Models (LAMs) that “predict” the next best step in a process, much like LLMs predict the next word. But there’s a catch—LAMs are only as powerful as the actions they've seen. They rely on data libraries to guide them, which limits them to known scenarios. 3️⃣ Automating Roles, Not Just Tasks: We’re not talking about automating a single job step anymore; Agentic AI is aimed at automating entire roles. Think of a legal assistant who no longer needs to research case law because the AI handles it—freeing them up for nuanced, strategic work. But how far should we go? 4️⃣ The Risks: Sure, Agentic AI can boost efficiency and reduce human error. But as we lean more on it, we risk creating a workforce dependent on automation. Jobs could be fully automated, even when a human touch is needed. And smaller businesses? They could be left behind, lacking the vast datasets that large enterprises have to fuel these advanced systems. 5️⃣ Balance is Key: Just because we can automate, should we? Agentic AI brings enormous potential, but it also demands responsibility. We need frameworks that ensure transparency, avoid bias, and keep human oversight intact where it matters most. Agentic AI could be a game-changer—but it’s up to us to guide it in a direction that adds value without compromising human insight and ethics. So, what do you think? Is your business ready for Agentic AI? Are you? This is extracted from Alan Pelz-Sharpe's column in the latest issue of KMWorld, subscribe to them for a regular dose of Alan and to read the entire article (and other leaders in the industry). #AgenticAI #FutureOfWork #Automation #AI #Innovation #RPA

💡 The Shift to AI Agents: Prepare for Pay-Per-Use Pricing 💡 The free AI buffet is ending, and metered pricing is on the rise. Companies are moving beyond generative AI assistants like ChatGPT and Co-Pilot to AI agents that can make decisions and perform tasks autonomously. But here’s the catch: you’ll soon be paying for every interaction. Salesforce is leading the charge, with plans to charge $2 per conversation for its new AI agents. The shift toward consumption-based pricing is about to change the landscape for companies relying on AI. 🧠 Key Takeaways: AI agents are more powerful than generative assistants. They don’t just assist—they make decisions and follow workflows. ⚡Prepare for metered costs. Instead of flat fees, every AI-driven interaction could come with a price tag. ⚡Identify where AI agents bring real ROI. Not every use case will justify the cost, so choose wisely. ⚡Budgeting for AI will become more complex. Companies will need to plan for AI costs like they plan for cloud services. Is your organization ready to manage the new economics of AI? The shift from free trials to usage-based pricing is happening fast. This is a quick snippet from We Love Ugly Data Podcast, Series 3, Episode 9. Head over to our YouTube channel to subscribe. #AI #AIagents #Salesforce #genAI #DigitalEconomy #BusinessStrategy