New Post: Build an X++ AI Tool for Copilot Studio in Dynamics 365 F&O From Adrià Ariste Santacreu Learn how to wire real F&O business logic into Copilot Studio. This guide shows you how to build an X++ AI tool, set up security, auto/hand-register the Dataverse Custom API, configure inputs (including company context), create a topic, publish, and test—end to end. Click here to continue reading Build an X++ AI Tool for Copilot […] https://lnkd.in/ebDSxxmQ

Our recent webinar revealed our complete AI development process - the strategy, tools, and workflow behind building a working compliance platform in 3.5 hours. The audience saw everything: how we handle context engineering challenges, navigate AI limitations, and turn rough prototypes into working applications. This changes product validation completely. Instead of months of documentation and wireframes, you can put working software in front of clients much faster. Key insights from the session: Strategy and context matter more than coding speed New business models become viable when execution costs plummet Better judgment becomes critical when features are "free" to add Stay tuned as we're creating more valuable assets for you. Ready to build at AI speed? #AIBuilds #ProductDevelopment #DigitalInnovation #ProductStrategy #SoftwareDevelopment #BuildFaster #ThisIsStoked

🚀 Leveling Up My AI Agent Journey: From Research to Automation Over the past week, I’ve been diving deeper into building AI Agents and experimenting with real-world use cases using the OpenAI Agents SDK. Here’s what I’ve been working on: 🔹 Implemented a Deep Research Agent to perform structured web searches and generate concise summaries. 🔹 Explored hosted tools like WebSearchTool, FileSearchTool, and ComputerTool to extend agent capabilities. 🔹 Designed a Planner Agent that breaks down a query into multiple web searches for more comprehensive results. 🔹 Built an Email Agent (powered by SendGrid) that transforms agent outputs into clean HTML reports and sends them automatically. 🔹 Connected everything into an end-to-end pipeline: Search → Structured Output → Report → Automated Email Delivery. This was not just about writing code — it was about understanding how different agent components can collaborate to create intelligent workflows. Excited to keep pushing forward — next steps will be experimenting with multi-agent collaboration and expanding into more complex autonomous systems. #AIAgents #LLMEngineering #Automation #MachineLearning #AIResearch

AI & Automation Starter Pack If you’re looking to get started or level up in building AI-powered workflows, here’s a toolkit worth exploring: Workflows: Zapier | Make | n8n | Viasocket LLMs: OpenAI | Anthropic Chatbots: ChatBot.com | Botpress | LandBot | ChatBase | Voiceflow Voice Agents: Vapi | Retell | ElevenLabs Data & Ops: Airtable | Notion | Monday | Google Sheets Portfolio: Short Loom/Vimeo demos + Notion | Gamma

Over the past couple years, we’ve been quietly laying the foundation for something big. We knew AI would eventually play a major role in construction operations, but we didn’t want to build it just because everyone else was. We wanted to build something purposeful. Something that actually helps. Today, I’m excited to share the first step in that journey: 🚀 Otto, our new AI assistant, is now live inside Ontraccr. Otto is designed to make every user more capable, creative, and confident inside the platform. It’s trained on Ontraccr’s features, workflows, and best practices. It helps users build faster, automate smarter, and get more out of the tools they already have. But Otto is just the beginning. This is the first of many AI-powered tools we’ll be rolling out over the coming months—tools that won’t just answer your questions, but help you design, automate, and execute complex workflows in seconds. We're building a future where AI doesn't replace people—it amplifies them. Check out the launch post and video in the share below. Excited for what’s ahead. #AI #constructiontech #automation #ontraccr #nocode #workflowdesign #otto

Just stumbled upon Sim Studio, an open-source AI agent builder backed by Y Combinator, and I’m genuinely impressed. If you're into AI automation, this one stands out. It’s built around a visual, drag-and-drop interface that lets you design smart, multi-agent workflows without drowning in code. These agents can reason, collaborate, and make context-aware decisions. You’re not limited to basic "if this, then that" flows. Instead, think of autonomous agents that adapt and act. It integrates with major models like OpenAI, Anthropic, Mistral, and local setups via Ollama. Additionally, it integrates seamlessly with tools like Slack, Gmail, Pinecone, Supabase, and your custom APIs. Some real-world use cases: ╰┈➤ Customer support agents who escalate when needed and learn from chats ╰┈➤ BI agents that analyze data and trigger actions ╰┈➤ Content pipelines with research, writing, and editing handled by different agents ╰┈➤ Email automation that understands tone, urgency, and intent ╰┈➤ Smart data pipelines that don’t just move data; they act on it What I also like is the tech stack. It’s built with Next.js, Bun, PostgreSQL (with pgvector), Zustand, Tailwind, and more. Very developer-friendly. You can start locally with npx simstudio, deploy via Docker, try the hosted version at sim.ai, or self-host it fully. This really feels like the next step: moving from simple automations to intelligent orchestration. It could save AI-first teams months of backend work. Try it live: sim.ai Code: github.com/simstudioai/sim Docs: docs.simstudio.ai Are you exploring AI agents yet? It's time to move from linear workflows to ones that can reason and act. #AI #Automation #OpenSource #LLM #AgentWorkflows #SimStudio #Sim #AgenticAI #AIAgents #Agents

🤖 Most AI agents today are stuck in 𝗽𝗿𝗼𝘁𝗼𝘁𝘆𝗽𝗲 𝗺𝗼𝗱𝗲. They can’t easily tap into the tools and data that enterprises run on. That’s why 𝗠𝗼𝗱𝗲𝗹 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗣𝗿𝗼𝘁𝗼𝗰𝗼𝗹 (𝗠𝗖𝗣) is a game-changer: 🔗 It gives LLMs a standard, open-source way to connect with enterprise systems ⚡️ Unlocks real business use cases instead of toy demos 🌍 Opens the door to an ecosystem of AI-ready integrations 🎥 In my latest video with Kong Inc., I answer the question “What is MCP?” and share why it could be one of the most important building blocks for the future of AI agents. Full video below in comments ⬇️

Building AI voice agents doesn’t have to take days of developer time. Our prototyping platform reduces build time from hours to just one. In many cases, it empowers non-technical teams to test and refine voice agents themselves. Faster prototyping means faster trust, fewer risks and lower costs. Read how Solvd is helping companies move from idea to working prototype with confidence: https://hubs.la/Q03JRRJV0

I'm continually impressed by how far voice agents have come. They're complex, powerful and changing how we talk to tech.    Check out our article on how to build cost-effective and scalable agents. You'll be building voice agents to handle customer support (or mother-in-laws) in no time! 

🚀 Just wrapped up a fun experiment — building my first multi-agent workflow in n8n! As PMs, one of our most common (and sometimes painful) processes is taking a product from Ideation → Prioritization. I decided to automate it using a chain of AI agents (powered by Gemini 2.5) connected through n8n. Here’s how it works: ✅ Ideas go into Google Sheets ✅ Each agent refines, processes, and prepares inputs for the next one ✅ Final prioritized output lands back in Google Sheets Along the way, I got hands-on with Google Sheets integration, a bit of JavaScript, and most importantly — the power of workflow automation with n8n. 📹 I’ve put together a quick video of the workflow in action. Would love to hear your thoughts and ideas on how you’d use multi-agent workflows! #ProductManagement #AI #WorkflowAutomation #n8n #GeminiAI