MAS Use Cases in 2025

How Are People Really Using #GenAI in 2025? 100 use cases. 38 new entries. 3 dominant shifts. Marc Zao-Sanders’ recent Harvard Business Review article presents a data-rich exploration of how users are applying generative AI in practice, based on Reddit, Quora, and other public forums. If this is true, it signals a decisive transition: from technical augmentation to cognitive and emotional co-processing. Top Use Cases by Impact (2025): 1. AI Therapy & Companionship Supporting grief processing, daily emotional regulation, and self-guided reflection, especially in regions with limited access to mental health resources. 2. Life Organization & Executive Function Support Assisting with timeline planning, behavioral nudging, and micro-habit scaffolding. This suggests a growing role for LLMs as external cognitive frameworks. 3. Purpose Discovery & Values Clarification Enabling identity shaping, life decision support, and long-term goal-setting through reflective, conversational interactions. ⸻ Key Trends: • From Instruction to Alignment LLMs are increasingly prompted not just for answers but for structure, motivation, and agency. • Prompt Engineering Maturity Users now craft multi-layered prompts with embedded objectives, constraints, and tone, often in collaboration with the model. • Tension Between Privacy and Personalization Users express both concern about data capture and frustration over limited memory retention. This highlights the paradox of context-aware systems. ⸻ If this shift holds at scale, it redefines the role of #LLMs. They are no longer just productivity tools but distributed cognitive collaborators that scaffold emotional reasoning, identity exploration, and behavioral planning. Enterprise design must evolve accordingly. We must support agentic models not just with #APIs but with psychological #intentionality, robust #transparency controls, and adaptive #context management. It is the era of AI-in-the-loop.

Let's be honest – most business default to MySQL or MongoDB without considering the broader database ecosystem. In 2025, this approach could be seriously limiting your applications, especially with AI workloads. Here are 12 database types that power modern tech: 1. Blockchain Databases    • Distributed ledger technology    • Smart contract integration    • Use cases: DeFi, NFTs, Supply Chain    • Examples: BigchainDB, Amazon QLDB 2. SQL Databases    • ACID compliance    • Complex joins & transactions    • Use cases: Financial systems, ERP    • Examples: PostgreSQL, MySQL, Oracle 3. Columnar Databases    • Column-based storage    • Data warehouse optimization    • Use cases: Business intelligence, analytics    • Examples: ClickHouse, Amazon Redshift 4. NewSQL Databases    • Horizontal scalability    • ACID guarantees    • Use cases: High-throughput applications    • Examples: CockroachDB, Google Spanner 5. In-Memory Databases    • Sub-millisecond latency    • Real-time processing    • Use cases: Caching, session management    • Examples: Redis, Memcached 6. Spatial Databases    • Geographic data optimization    • Spatial indexing    • Use cases: Maps, location services    • Examples: PostGIS, MongoDB Atlas 7. Vector Databases    • Similarity search    • Embedding storage    • Use cases: AI applications, recommendation systems    • Examples: Milvus, Pinecone 8. Graph Databases    • Relationship-first design    • Network analysis    • Use cases: Social networks, fraud detection    • Examples: Neo4j, Amazon Neptune 9. Time-Series Databases    • Time-based data optimization    • Efficient aggregation    • Use cases: IoT, monitoring    • Examples: InfluxDB, TimescaleDB 10. Key-Value Databases    • Simple data model    • High throughput    • Use cases: Shopping carts, user sessions    • Examples: DynamoDB, etcd 11. Document Databases    • Schema flexibility    • JSON/BSON storage    • Use cases: Content management, catalogs    • Examples: MongoDB, Couchbase 12. Object-Oriented Databases    • Object persistence    • Complex data structures    • Use cases: CAD/CAM, scientific data    • Examples: ObjectDB, db4o Key Architectural Considerations: Performance: Query optimization patterns Indexing strategies Caching mechanisms Load balancing approaches Scalability: Vertical vs horizontal scaling Consistency models Partitioning strategies Replication mechanisms Production Readiness: Monitoring & observability Backup & disaster recovery Security best practices Cost optimization AI Integration: • Vector search implementation • Embedding storage patterns • Real-time inference support • Training data management Real-world Considerations: • Cost optimization strategies • Monitoring & observability • Backup & disaster recovery • Security best practices

Most AI teams are burning $100K/month because they're using the wrong model for the job. After analyzing 600+ enterprise AI implementations, I discovered why 73% of AI projects fail to scale. The secret? They're force-fitting one model architecture when they need specialized intelligence. The 8 Specialized AI Models That Will Define 2025's Winners (Visual guide attached 👇) → LLMs (Large Language Models) The workhorses. But using GPT-4 for everything is like using a Ferrari for grocery runs. Production tip: Token-by-token processing enables incredible reasoning but costs explode without proper orchestration. → LAMs (Large Action Models) The game-changer nobody's talking about. These don't just think—they execute. Security note: LAMs require robust sandboxing. We've seen breaches from improper action boundaries. → LCMs (Large Concept Models) Meta's revolutionary approach encoding entire sentences as concepts. Why this matters: 40% faster inference, 60% less compute. → MoE (Mixture of Experts) Activate only what you need. Like having specialist consultants on-demand. Cost insight: Reduces compute by 78% while maintaining GPT-4 performance. → VLMs (Vision-Language Models) See + understand + reason. The backbone of next-gen automation. Compliance critical: GDPR requires explainable visual AI decisions. → SLMs (Small Language Models) David vs Goliath. Powering the edge AI revolution. Security advantage: On-device processing = zero data leakage. → MLMs (Masked Language Models) The OG bidirectional champions. Still unbeatable for certain tasks. Use case: Financial compliance requires context from both directions. → SAMs (Segment Anything Models) Pixel-perfect precision. Foundation of visual AI automation. ROI metric: One SAM deployment replaced 12 manual QA engineers. Here's what kills most AI projects: Using one architecture for everything. The winning formula across 50+ enterprises: Map use case to right architecture Build security-first from day one Implement responsible AI guardrails Monitor costs per inference Design for human-AI collaboration The paradigm shift: Stop asking "Which AI model should we use?" Start asking "Which specialized architecture solves this specific problem?" This is why vibe coding with the right model beats traditional development every time. Your competitive edge in 2025: Companies matching specialized architectures to tasks are seeing: ↳ 10x faster deployment ↳ 80% cost reduction ↳ 99.9% uptime with proper orchestration ↳ Zero compliance violations We help enterprises implement these architectures with bank-grade security and SOC2 compliance baked in. 💡 Power insight: The future isn't one super-intelligent AI. It's orchestrated specialist AIs working in harmony. What specialized AI architecture could transform YOUR biggest bottleneck? Drop your use case below and I'll personally recommend the right architecture ⬇️

🌍 2025 The Year the #Payment #Card #Lost Its #Center Stage For the past two decades, Visa and Mastercard have been the global synonyms for cashless #payments 💳, controlling over 85% of the world’s #transaction volume. Their rails were the standard the entire industry relied on. 💡 The shift: In 2015, “new #money” lived alongside cards. By 2025, it’s aiming right at the core of payment infrastructure. And it’s not just fintech startups 🚀 but also regulators 🏛️ - the very players who seemed to be card allies not long ago. #Visa and #Mastercard remain strong - $18 trillion in annual volume, infrastructure in 200+ countries 🌐. But for the first time in 40 years, they face a challenger that’s not just “another way to pay” but a force changing the architecture 💥. 📈 2025 facts: • 🇪🇺 In Europe, #A2A share in e-commerce has passed 20% • 🇳🇱🇵🇱 In the Netherlands and Poland - over 40% • 🇮🇳 India’s #UPI processes 75% of P2P and is expanding offline • 🇧🇷 Brazil’s #Pix closed 39B transactions in 2024 and is still growing at double digits • 💲 Circulating #stablecoins have surpassed $160B, with #USDC and #USDT accepted offline from 🇹🇭 #Thailand to 🇭🇰 Hong Kong 📌 Case #1 - Colombia: pharmacy chain Droguerías Colsubsidio enabled USDC payments via a local #PSP directly in #POS terminals. Average cashless fees dropped from 2.5% to 0.4%, with settlement times down to under a minute ⏱️. 📌 Case #2 - Europe: In January 2025 🇳🇱 marketplace bol.com announced that within 6 months, 32% of all payments had shifted to A2A, with direct stablecoin settlements for cross-border sellers. Settlement times fell from T+2 to T+0, and refund costs dropped 5x 📉. 💡 Why it matters: A2A connects bank-to-bank 🏦↔🏦, cutting out intermediaries and fees. Stablecoins enable instant cross-border payments 24/7 🌐⚡. For consumers - lower costs and faster payments. For businesses - better margins and new markets 📊. 🔮 Outlook: If current growth continues, by 2030 up to 50% of online transactions in developed markets could bypass card networks entirely. The irony: cards are too good to die fast 💳✨. But too outdated to win the long game 🏁.

MAG (Airports Group) built an Agentic AI system to solve unplanned staff absences across their airport operations, in collaboration with the AWS Generative AI Innovation Center.    The use case is straightforward but the complexity makes it ideal for Agentic AI to solve: When an employee calls in sick at an airport running over 1000 flights daily, someone needs to authenticate the employee, apply the correct HR policy based on absence type, update multiple systems, notify managers, and re-roster shifts. Different airports have different job types and different absence categories, so there are hundreds of workflow permutations.   MAG built an Agentic AI solution using Amazon Bedrock AgentCore. The system conducts natural conversations with employees (using Amazon Nova Sonic), applies context-aware guardrails for compliance with critical infrastructure requirements, and coordinates across multiple backend systems. They achieved 99% consistency in absence reporting and reduced processing time by 90%.   The video from AWS re:Invent 2025 is worth watching to understand the technical approach and lessons learned. They used multi-agent architecture where a speech-to-speech model connects to a text-based agent that orchestrates the tools, built custom guardrails that track conversation patterns across multiple turns rather than just screening individual messages, and separated tools from agents using model context protocol for reusability.   https://lnkd.in/e4NiQKVW This is the first use case in MAG's vision for a digital colleague workplace where multiple AI agents coordinate airport operations, with the full session covering their security implementation for critical national infrastructure and practical lessons on adapting tools for agents and maintaining user engagement through thoughtful interface design.   The complexity inherent in airport operations is exactly where Agentic AI demonstrates its value over traditional automation. Thanks Steve Campagnaro Margherita Rosnati Tom Chester for your insightful presentation and innovative solution! Amazon Web Services (AWS) #airportAI

Here is the AI use case & implementation report that I’ve built from 25+ conference talks by supply chain leaders. The reason? Leaders are asking: "How are my peers actually implementing AI and what is just marketing vs. actual results?" Based on the 10+ conferences I attended this year, I've synthesized 25+ of the most practical talks with direct links to the recordings and slides into a consolidated view of what is working now. No sales pitches, no theory, just VP and Director-level practitioners sharing their wins and losses. This report is your map to the practical solutions others are successfully implementing, enabling you and your team to navigate your own AI journey with greater confidence. Inside, you'll find: 1️⃣ Key patterns emerging in 2025, including the shift from "hype to measurable ROI" and the adoption of "agentic automation". 2️⃣ Actionable "how to apply" sections and specific case studies from leaders at The Hershey Company, Toyota Motor Corporation, Renault Group, Bayer, and A.P. Moller - Maersk. 3️⃣ Breakdowns by supply chain function: including AI implementation & strategy, procurement, planning & visibility, and transportation & fulfillment. 4️⃣ Direct links to the 25+ original talks analyzed for this report. A huge thank you to all the speakers and leaders who are sharing their knowledge so openly. It is this spirit of collaboration that will move our entire industry forward. Lori Boyer, Tony Filippone, Matthew Barry, Douglas Guilherme, Niraj Jha, Sean Jacobsohn, Shashi Mandapaty, Dr. Elouise Epstein 🏳️⚧️, Rosalia Snyder, Bawana Radhakrishnan, Jeanne C M., Eva Choe, Al Williams, Sebastiano Finocchiaro, Paul Polman, Janelle Aydin, Bertrand Conquéret, Mithun Sharma, Felix M., Maria Jesus Saenz, Scott Gaston, Arne Jeroschewski, Tamer Al Ghussein, Pavel Sinelnik, Tiago Paiva, Swagat Choudhury, Craig Sutton, Michael Castagnetto, James Lamont, Jean-Marc Carlicchi, Fabian Pobantz, David Herb Download the full report using the link in the comments.

In 2025, AI Agents will be everywhere. Only a few will actually save you money. What are the most common 𝗔𝗜 𝗔𝗚𝗘𝗡𝗧 𝘂𝘀𝗲 𝗰𝗮𝘀𝗲𝘀? → Agentic RAG: They retrieve knowledge data, evaluate sources, reason, and deliver contextually grounded answers. Perfect for internal knowledge assistants or enterprise Q&A. Examples: IBM Watsonx, Glean. → Workflow Automation Agents: Trigger tasks across systems without human involvement. Think onboarding flows or approvals. Examples: Make, n8n, Zapier. → Coding Agents: These agents can plan, refactor, debug, and even reason across repositories. Not just code suggestions. Examples: Cursor, Claude Code, Copilot. → Tool-Based Agents: Designed for specific tools and defined tasks like lead enrichment or sending emails. Examples: Breeze, Clay, Apollo. → Computer Use Agents: They navigate UIs like humans: clicking buttons, typing forms, and browsing. Powered by models like Claude and GPT. → Voice Agents: Handle calls for support, sales, or internal queries. All with voice Examples: Retell AI, Vapi AI. AI Agents are reshaping workflows, but only if you use the right ones. Which of these use cases are you exploring in your organization? Share your thoughts!

AI Agents Are No Longer Just Hype — They're Powering Real-World Use Cases! From automating workflows to coding apps 10x faster, AI agents are transforming how we work, build, and interact with technology. Here's a breakdown of the most popular use cases dominating the AI landscape in 2025: Voice Agents Used in customer service with Speech-to-Text (STT), Text-to-Speech (TTS), and telephony integrations. Think virtual call agents powered by tools like Vapi and ElevenLabs. Agentic RAG (Retrieval-Augmented Generation) These agents combine retrieval with generation for smarter answers. Powered by tools like Perplexity, Glean, and vector databases such as Pinecone or Weaviate. Workflow Automation Platforms like n8n, FlowiseAI help users create agents that manage systems, internal APIs, and services like Gmail or Stripe—automating repetitive business logic effortlessly. Computer User Agents Versatile agents that control browsers, editors, and apps via the UI. Think GPT-based copilots that simulate human interaction with your computer. Coding Agents Agents like Cursor and Roo Code that write, debug, and understand code—boosting developer productivity with multi-agent collaboration. Tool-Based Agents Highly specialized agents that work with specific tool stacks (e.g., Kogi, Clay, Breez) by connecting to dedicated APIs across services. These use-cases highlight the versatility and modular power of modern AI agents. Whether you're an engineer, startup founder, or enterprise leader — there's likely an agent that can save you time, reduce friction, and supercharge your workflows. Which use case do you think will have the biggest impact in the next 12 months?

𝗗𝗥𝗘𝗔𝗠𝗙𝗢𝗥𝗖𝗘 𝟮𝟬𝟮𝟱 - 𝗜‘𝗺 𝗶𝗻 𝗦𝗮𝗻 𝗙𝗿𝗮𝗻𝗰𝗶𝘀𝗰𝗼 𝘄𝗶𝘁𝗵 𝗦𝗮𝗹𝗲𝘀𝗳𝗼𝗿𝗰𝗲!! ☁️💙 [𝗔𝗱/𝗔𝗻𝘇𝗲𝗶𝗴𝗲] Many companies like to talk about AI painting big pictures of what might come next. Salesforce takes a different approach: they build! For more than a year now, Salesforce has been rolling out AI agents that are already running inside companies around the world. In yesterday’s keynote, 𝗠𝗮𝗿𝗰 𝗕𝗲𝗻𝗶𝗼𝗳𝗳, 𝗖𝗘𝗢 𝗼𝗳 𝗦𝗮𝗹𝗲𝘀𝗳𝗼𝗿𝗰𝗲, shared new use cases that made it easier than ever to understand how these agents really work in daily operations. 𝗕𝘂𝘁 𝗼𝗻𝗰𝗲 𝗮𝗴𝗮𝗶𝗻: 𝗪𝗵𝗮𝘁 𝗶𝘀 𝗮𝗻 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁? Imagine opening your laptop and finding a small team of digital helpers already inside. Each one is an expert… one knows your customers very well, one your workflows, one is your data expert. They don’t just answer your questions or react to commands but fix things before you see them and make work feel more fluent, faster, personal & fun. Marc Benioff described this evolution clearly: “𝘈 𝘺𝘦𝘢𝘳 𝘢𝘨𝘰, 𝘈𝘨𝘦𝘯𝘵𝘧𝘰𝘳𝘤𝘦 𝘸𝘢𝘴 𝘢 𝘱𝘳𝘰𝘥𝘶𝘤𝘵. 𝘛𝘰𝘥𝘢𝘺, 𝘪𝘵’𝘴 𝘵𝘩𝘦 𝘱𝘭𝘢𝘵𝘧𝘰𝘳𝘮 𝘣𝘦𝘩𝘪𝘯𝘥 𝘦𝘷𝘦𝘳𝘺𝘵𝘩𝘪𝘯𝘨 𝘸𝘦 𝘥𝘰.” 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗳𝗼𝗿 𝗰𝗼𝗺𝗽𝗮𝗻𝗶𝗲𝘀: –> up to 30 % faster service resolution and 40 % lower response time in customer operations –> productivity gains between 20–35 % across early adopters –> now used by 12,000+ companies from retailers to logistics firms –> interoperability with AWS, Microsoft & OpenAI, so it fits into existing tech stacks –> built-in governance and transparency layers, critical for regulated industries 𝗟𝗲𝘁’𝘀 𝘁𝗮𝗹𝗸 𝗮𝗯𝗼𝘂𝘁 𝘀𝗼𝗺𝗲 𝗰𝗼𝗻𝗰𝗿𝗲𝘁𝗲 𝘂𝘀𝗲 𝗰𝗮𝘀𝗲𝘀: 𝗪𝗶𝗹𝗹𝗶𝗮𝗺𝘀 𝗦𝗼𝗻𝗼𝗺𝗮 – An AI “shopping chef” that knows your taste. It connects recipes, products, and past purchases turning every visit into a personalized experience that feels more like a conversation than a normal shopping experience. 𝗙𝗲𝗱𝗘𝘅 – AI agents read and route thousands of logistics documents in seconds catching exceptions, rerouting shipments, and reducing manual work across global operations. 𝗣𝗮𝗻𝗱𝗼𝗿𝗮 – An AI assistant follows you from online to in-store. What you like in chat appears ready in the boutique creating one seamless, personalized customer journey. 𝗗𝗲𝗹𝗹 – AI agents automate supplier onboarding verifying documents, sending approvals, and cutting setup time from 60 days to under 20. Faster partnerships, faster production. Each example shows how AI can move beyond experimentation, into real outcomes & that‘s what we need more now: REAL IMPLEMENTATION! Tomorrow continues with more 𝗧𝗲𝗰𝗵 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 𝗳𝗿𝗼𝗺 𝗮 𝗚𝗟𝗢𝗕𝗔𝗟 𝗦𝗨𝗣𝗘𝗥𝗦𝗧𝗔𝗥 I was able to meet and we all probably know more for his music than for his tech… STAY TUNED!! 💙🦾 Do you already use AI Agents in YOUR business? –> If yes, what for? –> If not, which tasks would you 𝘭𝘰𝘷𝘦 to hand over to an agent friend?

📢 Quarterly Mapping: The Re-Use of Non-Traditional Data in Practice 👉 Every quarter, we map and share how non-traditional data—mobility traces, platform data, wastewater signals, crowdsourcing—is being used for public-interest applications. Our Q4 2025 mapping (Sept–Dec) includes several use cases related to: • Public health intelligence & early warning • Crisis response & humanitarian decision-making • Climate resilience, air quality & infrastructure damage • Urban mobility, labor markets & political sentiment • Migration and conflict-affected settings 🔎 The big takeaway? 📍 Non-traditional data is no longer experimental. 📍 It’s being used to complement official data in terms of speed and granularity. 📍 But it’s still rarely institutionalized 📉 The key challenge remains governance: Who gets access? Under what conditions? And how do we move from ad-hoc reuse to trusted, durable pathways for public-interest data use? 📄 Full quarterly overview here: Recent Uses of Non-Traditional Data in the Public Interest (✍️ with Adam Zable): https://lnkd.in/ePnDTdEk #NonTraditionalData #PublicInterestData #DataForPolicy #EvidenceInAction #DataGovernance #CrisisAnalytics