Design

LLM for Spatial Understanding – SpatialLM on Hugging Face The model has been around as a research project, but now it’s ready to use - pretrained, open-source, and much more accessible. What’s interesting is how it actually works, and what “understanding” means in this context: 𝐒𝐩𝐚𝐭𝐢𝐚𝐥𝐋𝐌 takes visual input (like a simple phone video), reconstructs it into a 3D point cloud, and then passes that through an LLM. But the LLM doesn’t just process words, it uses its semantic knowledge to interpret geometry. So instead of saying “Here’s a flat shape,” it says: “That’s a wall. There’s a door attached to it. That’s a sofa, facing this direction, with these dimensions.” And it doesn’t stop there. The output is structured, machine-readable data. For example: Bbox = Bbox(“sofa”, position=(2.9,1.6,3.7), size=(1.7,0.8,1.8)) This kind of fusion really excites me: 𝐯𝐢𝐬𝐢𝐨𝐧 + 𝐥𝐚𝐧𝐠𝐮𝐚𝐠𝐞 + 𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 Not just detecting objects, but reasoning about space. Not just pixels, but meaning. I can see this unlocking smarter AR, robotics, indoor mapping, and more. And it’s open-source, built on LLaMA and Quinn, trained on real-world video. Definitely one to keep an eye on! 📍Btw, we also just open-sourced 𝐆𝐞𝐧𝐀𝐈 𝐀𝐠𝐞𝐧𝐭𝐎𝐒, a lightweight framework we’ve been using internally to run multi-agent systems. If you’re playing around with agent workflows, feel free to check it out: GitHub: https://bit.ly/4kzE1Mt And if you’re into open source, a ⭐ would mean a lot! __________ For more on AI and open source, plz check my previous posts. I share my journey here. Join me and let's grow together. Alex Wang #generativeai #ai #aiagents #llms #opensource

When packaging becomes part of you Wearable packaging is no longer a futuristic concept it's a growing design frontier that merges functionality, fashion, and emotional connection. Beyond the shelf: packaging you can wear From fragrance necklaces to ring-shaped lip balms and refillable compacts designed like jewelry, beauty brands are transforming packaging into accessories. The product becomes not just something you use, but something you wear turning daily rituals into statements of identity. Why it works In a world saturated with options, wearable packaging offers immediacy, memorability, and emotional value. It adds a layer of meaning: a perfume worn around the neck is not just a scent, it’s a story you carry. A lipstick shaped like a pendant becomes both utility and ornament. It taps into consumers’ desire for customization, portability, and aesthetic expression particularly among Gen Z and Millennials who seek objects that are both functional and symbolic. A new layer of storytelling Wearable formats create a deeper connection between brand and user. They’re conversation starters, collectible, and often more sustainable built for reuse and ritual, rather than discard. Design becomes not only a visual differentiator but a tactile, emotional experience that enhances the perceived value of the product. From beauty to fashion and back again This blurring of categories also opens doors to cross-industry innovation. When beauty products act like accessories, they enter the realm of fashion making room for brand collaborations, limited editions, and viral potential on social media. It’s no longer just about packaging design. It’s about presence. Relevance. And creating a product people don’t want to put away — they want to put on. Featured brands: Rhode Bubble Yepoda Laneige #WearablePackaging #CosmeticAccessories #FragranceOnTheGo #LuxuryPackaging #EmotionalDesign #GenZBeauty #FutureOfBeauty

The PayPal, Mastercard and Visa announcements are not about agentic AI. They are about ownership of the next chapter of commerce and payments. Here is how they compare. 𝗪𝗵𝗮𝘁 𝗵𝗮𝘀 𝗯𝗲𝗲𝗻 𝗮𝗻𝗻𝗼𝘂𝗻𝗰𝗲𝗱: -   PayPal : APIs that let any AI agent pay, track shipping, issue invoices and resolve disputes without leaving the chat. -   Mastercard : Network tokens + passkeys so agents become “trusted purchasers,” with programmable rules and biometric SCA baked in. -   Visa:  Five modular APIs for discovery → checkout, including user‑set spend caps, MCC filters and real‑time approvals. 𝗪𝗵𝗮𝘁 𝗶𝘀 𝗮𝘁 𝘀𝘁𝗮𝗸𝗲? -   The payments race has always been about shaving seconds off checkout. In the agentic era, the winning time is 0 seconds, 0 clicks. Checkout disappears entirely as search, recommendation, and payment collapse into a single LLM-driven conversation. -   Whoever owns the payment credential becomes the default wallet in the loop, capturing not just the transaction, but data, interchange, and value-added services that follow. -   The players that get this right won’t just win conversions. They’ll own the customer relationship. The ones that don’t will find themselves disintermediated by someone else’s agent. 𝗧𝗵𝗲 𝗽𝗼𝘁𝗲𝗻𝘁𝗶𝗮𝗹: Imagine: • A travel bot books flights, hotels, insurance and pays - no forms. • An SME sourcing agent negotiates fabric in Guangzhou and settles with a virtual card - no emails. • A grocery assistant notices the fridge is low and re‑orders - no conscious decision. Multiply that by every vertical and every consumer. That’s always‑on demand capture - and potentially trillions in incremental volumes routed through whoever provides the agent‑native rails. 𝗪𝗵𝗮𝘁’𝘀 𝗹𝗶𝗸𝗲𝗹𝘆 𝗻𝗲𝘅𝘁: 1.     Industry standards: Common schemas for trusted-agent registration, permissions, and dispute handling. 2.     Granular consumer controls: Per-transaction biometrics, spend limits, time-of-day and merchant-category restrictions. 3.     Merchant enablement: SDKs and APIs to expose real-time inventory, pricing, and loyalty programs to agents. 4.     Regulatory attention: How frameworks like PSD3, CFPB guidelines, or MAS oversight will apply to autonomous payers. 5.     New revenue models: Pay-per-call risk scoring, agent onboarding fees, premium fraud protection layers. 6.     Advanced risk infrastructure: Real-time monitoring of agent behaviour, intent detection, and adaptive risk scoring to flag anomalies. 7.     Liability frameworks: Clear rules for who’s accountable when an agent transacts incorrectly: the user, the platform, or the agent provider. The race to build the payment infrastructure for autonomous agents is underway. Expect a wave of partnerships, acquisitions, and early execution challenges as the industry adapts to a new model of always-on, agent-driven commerce. Opinions: my own 𝐒𝐮𝐛𝐬𝐜𝐫𝐢𝐛𝐞 𝐭𝐨 𝐦𝐲 𝐧𝐞𝐰𝐬𝐥𝐞𝐭𝐭𝐞𝐫: https://lnkd.in/dkqhnxdg

Supply Chain Sustainability 🌎 According to Deloitte, advancing supply chain sustainability requires embedding four strategic dimensions—net zero, resilience, compliance, and circularity—across five critical stages: risk assessment, strategy definition, stakeholder engagement, implementation, and ongoing performance management. This framework enables companies to align operational practices with emerging regulatory expectations and long-term value creation. The assessment phase focuses on identifying regulatory trends, emissions baselines, supply chain visibility, and data integrity gaps. These diagnostics support the quantification of risk exposure and inform the development of a maturity baseline, enabling more precise target-setting and investment prioritization. Strategy formulation involves translating insights into actionable standards and governance. This includes setting sustainability KPIs for procurement categories, reviewing supplier codes of conduct, and establishing cross-functional steering mechanisms. These elements provide the structure required to operationalize sustainability at scale across global supply networks. Implementation extends to integrating sustainability into product design, procurement, and logistics. Deloitte highlights initiatives such as renewable energy sourcing, clean transport deployment, and resource-efficient packaging. Embedding ESG and EHS standards into operational processes reinforces compliance and reduces exposure to reputational and operational risks. The final stage—monitoring and management—ensures continuous alignment through due diligence protocols, contract integration, supplier audits, and real-time risk sensing. These mechanisms not only track performance but enable agile responses to disruptions, making sustainability a core driver of supply chain resilience and competitive advantage. #sustainability #sustainable #business #esg

India has 150 million+ people above the age 60 and there is a massive opportunity to keep them healthy & fit. But everyone’s focused on Gen Z and no one’s building for their parents. It’s a hard business but a big one. We’ve invested in two companies. Here’s why it’s tough and how one should crack it. Understand the reality first. 1. Elders don’t think of “health” as proactive. They’re conditioned to wait until something breaks before acting. You’re selling a solution to a problem they don’t know they have yet. 2. The 65-year-old needs it but their 35-year-old child pays for it. You're not selling to the elder. You’re selling to their guilt-driven kids in Gurgaon or US. The buyer ≠ the user. 3. Trust is everything and you don’t have it. Indian elders trust: Their doctor, astrologer & their neighbour Not apps. Not tech bros. Not AI. You can't growth hack trust. You earn it slowly, locally. 4. They don’t want new habits. They’ve had the same breakfast for 40 years. You’re not selling a product. You’re undoing decades of routine. 5. Distribution is hyperlocal. Elders don’t click Insta ads. They talk to the uncle in their colony. You scale building by building not by user cohorts. Yes, 150M+ elders. But it’s not one market. It’s a thousand tiny tribes. Different languages, cultures, food habits, family structures, and tech comfort levels. If it were easy, Tata or Reliance would’ve done it already. But it’s wide open now. The one who combines tech + trust + real care will win. So how do you crack it? 1. Think first principles & not trends Don’t build a “senior fitness app.” Ask: Why did they stop moving? What gives them joy? You’re selling independence, not health. 2. Design for peace, not features. One-click help, One daily routine, One trusted face. Great elder products feel like human care not software. 3. Human-first, tech-enable. Don’t replace the daughter. Support her. Train 100 amazing elder coaches. Build tools to help them scale. 4. Don't focus on CAC. Here, it’s about trust per acquisition. You’re not selling toothpaste. You’re asking to be let into their daily life. Start offline. Build trust then tech. 5. You’re in the business of habit change & not selling an app or a pill. Get them to walk 15 minutes a day. Add protein to breakfast. Laugh more. Sleep better. Small wins compound. Don’t build for scale first. Build for consistency. Be in the business of habit change. 6. This isn’t a hackable D2C play. It’s a decade-long trust business. Build for one community. Get to know 100 elders by name. Solve deep, boring problems with elegance. Everyone’s chasing the next billion youth users. But the hidden opportunity lies in serving the first 150 million elders. The elder care market in India isn’t just underserved. It’s misunderstood and needs long-term play. Founders who crack this will build generational companies.

𝗡𝗼𝘁 𝗮𝗹𝗹 𝗔𝗣𝗜𝘀 𝗮𝗿𝗲 𝗰𝗿𝗲𝗮𝘁𝗲𝗱 𝗲𝗾𝘂𝗮𝗹. The architecture you choose has a direct impact on your application's scalability, performance, and long-term maintainability. It's not just about implementation—it's about alignment with your business goals and technical constraints. Here’s a breakdown of six core API architectures every engineer and architect should be familiar with: 𝟭. 𝗥𝗘𝗦𝗧 (𝗥𝗲𝗽𝗿𝗲𝘀𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗦𝘁𝗮𝘁𝗲 𝗧𝗿𝗮𝗻𝘀𝗳𝗲𝗿)   Best suited for simple, resource-oriented applications.   REST is stateless and leverages standard HTTP methods (GET, POST, PUT, DELETE). It’s easy to implement, scalable, and well-suited for CRUD operations in web apps. 𝟮. 𝗚𝗿𝗮𝗽𝗵𝗤𝗟   Ideal for applications that need flexible, efficient data fetching.   Unlike REST, GraphQL lets clients specify exactly what data they need, reducing both over-fetching and under-fetching. It’s a strong choice for front-end-heavy applications with complex or evolving data needs. 𝟯. 𝗦𝗢𝗔𝗣 (𝗦𝗶𝗺𝗽𝗹𝗲 𝗢𝗯𝗷𝗲𝗰𝘁 𝗔𝗰𝗰𝗲𝘀𝘀 𝗣𝗿𝗼𝘁𝗼𝗰𝗼𝗹)   Best for high-security, enterprise-grade applications.   SOAP is XML-based, highly structured, and supports strict security and compliance standards. It’s commonly used in sectors like banking and healthcare where data integrity and reliability are non-negotiable. 𝟰. 𝗴𝗥𝗣𝗖 (𝗚𝗼𝗼𝗴𝗹𝗲 𝗥𝗲𝗺𝗼𝘁𝗲 𝗣𝗿𝗼𝗰𝗲𝗱𝘂𝗿𝗲 𝗖𝗮𝗹𝗹)   Designed for high-performance, low-latency distributed systems.   Using HTTP/2 and Protocol Buffers, gRPC supports fast, compact communication and bidirectional streaming—making it a powerful choice for microservices and mobile apps. 𝟱. 𝗪𝗲𝗯𝗦𝗼𝗰𝗸𝗲𝘁𝘀   Best for real-time, interactive applications.   WebSockets establish persistent, two-way communication between client and server. This is essential for use cases like chat systems, live notifications, and multiplayer gaming. 𝟲. 𝗠𝗤𝗧𝗧 (𝗠𝗲𝘀𝘀𝗮𝗴𝗲 𝗤𝘂𝗲𝘂𝗶𝗻𝗴 𝗧𝗲𝗹𝗲𝗺𝗲𝘁𝗿𝘆 𝗧𝗿𝗮𝗻𝘀𝗽𝗼𝗿𝘁)   Optimized for IoT and sensor-based environments.   MQTT is a lightweight protocol designed for low-bandwidth, power-constrained devices. Its publish-subscribe model is ideal for real-time telemetry and remote monitoring. 𝗪𝗵𝗲𝗻 𝘁𝗼 𝘂𝘀𝗲 𝘄𝗵𝗮𝘁?   - REST: General-purpose web apps   - GraphQL: Dynamic queries, flexible UIs   - SOAP: Secure enterprise systems   - gRPC: Microservices, real-time communication   - WebSockets: Live, event-driven apps   - MQTT: IoT, embedded devices The right API architecture doesn’t just support your product—it accelerates it. It improves performance, simplifies integration, and positions your application for scale. What API design do you rely on the most? And what lessons have you learned from working with it? Gif Credit - Nelson Djalo

One of the easiest ways to improve your collaboration with product managers and engineers is to improve your design handoff files. - It doesn’t take a lot of time - It drastically improves how people interpret your designs - It reduces backs and forths between engineers and designers - It gives engineers confidence that they are doing the right thing - It increases the chances that what’s implemented will match the design Great handoff files, for me, are an instant sign of design maturity. It shows me that designers think not only of themselves but also the ecosystem of people around them. Coming up with great handoffs boils down to the following: - Setting context - Adding structure - Adding annotations - Including all states - Visualizing the flow - Including a prototype - Doing a run-through Not everything is required for every design initiative; a small feature update may not need a full handoff, whereas a big, impactful one may require everything from the above. In this cheat sheet, I’ve put together my top tips for delivering the “perfect” handoff file. Bonus: I have also created a Figma Annotation and handoff kit with handy components for the above. Find the link in the comments. 👇 — If you found this useful, consider reposting ♻️ PS: By handoff I'm not referring to the process of handing over a design, which should be a collaborative process from problem to solution, but rather the organized design file that is the single source of truth for what needs to be implemented. #uxdesign #uiux #productdesign

I interviewed 20 sustainability managers 🎙️ That's their #1 pain point 🤕 ➡️ "Reporting is 1st. Impact is 2nd". Challenges that I can see with sustainability in companies: ❌ Competing frameworks confuse. ❌ Data collection becomes more important than actual impact ❌ Disconnect between reporting teams and operational teams ❌ Excessive time spent on documentation. ❌ Risk of greenwashing through selective reporting (I am sure you have your observations to add🙄) 5 secrets to turn this into the biggest opportunity for change: ✅ Use reporting to clarify sustainability vision 100%. ✅ Identify in-company 'spoilers' - and engage them! ✅ Change sustainability reporting from 'a burden' for all, to an 'invitation to do good' for each individual. ✅ Turn deadlines into celebration moments for internal change. ✅ Use data requirements as opportunities to understand the entire value chain (and opportunities for change). You know the pain ?🧐 📲 Ping me to re-write the script on your sustainability reporting ♻️ #circulareconomy #zerowaste #sustainability

AI Product Management AI Product Management is evolving rapidly. The growth of generative AI and AI-based developer tools has created numerous opportunities to build AI applications. This is making it possible to build new kinds of things, which in turn is driving shifts in best practices in product management — the discipline of defining what to build to serve users — because what is possible to build has shifted. In this post, I’ll share some best practices I have noticed. Use concrete examples to specify AI products. Starting with a concrete idea helps teams gain speed. If a product manager (PM) proposes to build “a chatbot to answer banking inquiries that relate to user accounts,” this is a vague specification that leaves much to the imagination. For instance, should the chatbot answer questions only about account balances or also about interest rates, processes for initiating a wire transfer, and so on? But if the PM writes out a number (say, between 10 and 50) of concrete examples of conversations they’d like a chatbot to execute, the scope of their proposal becomes much clearer. Just as a machine learning algorithm needs training examples to learn from, an AI product development team needs concrete examples of what we want an AI system to do. In other words, the data is your PRD (product requirements document)! In a similar vein, if someone requests “a vision system to detect pedestrians outside our store,” it’s hard for a developer to understand the boundary conditions. Is the system expected to work at night? What is the range of permissible camera angles? Is it expected to detect pedestrians who appear in the image even though they’re 100m away? But if the PM collects a handful of pictures and annotates them with the desired output, the meaning of “detect pedestrians” becomes concrete. An engineer can assess if the specification is technically feasible and if so, build toward it. Initially, the data might be obtained via a one-off, scrappy process, such as the PM walking around taking pictures and annotating them. Eventually, the data mix will shift to real-word data collected by a system running in production. Using examples (such as inputs and desired outputs) to specify a product has been helpful for many years, but the explosion of possible AI applications is creating a need for more product managers to learn this practice. Assess technical feasibility of LLM-based applications by prompting. When a PM scopes out a potential AI application, whether the application can actually be built — that is, its technical feasibility — is a key criterion in deciding what to do next. For many ideas for LLM-based applications, it’s increasingly possible for a PM, who might not be a software engineer, to try prompting — or write just small amounts of code — to get an initial sense of feasibility. [Reached length limit. Full text: https://lnkd.in/gYY-hvHh ]

🧑🏼 How To Design Better Personas In UX (https://lnkd.in/eGPXmPNZ), a step-by-step guide to reduce decoration and add meaningful data to make personas more helpful and effective. Neatly put together by Slava Shestopalov. ✅ We need to know who users are and what they need to do. ✅ We can use both personas and Jobs-to-Be-Done for that. 🤔 They serve different purposes and focus on different things. ✅ Jobs-to-Be-Done focuses on user needs and outcomes. ✅ Personas focus on users, their behavior and mental model. ✅ Useful personas emerge from profound user research. ✅ They help visualize users, their goals and motivation. 🚫 Don’t focus on demographics to avoid stereotypes. ✅ Include the way of thinking, background, “a day in life”. ✅ Always add at least one persona with a disability. ✅ Add a story, pain points and how they use your product. ✅ List user’s habits/products they use daily, often and rarely. ✅ Finally, add needs, wants and fears mentioned by users. ✅ Then, prioritize key points for each role in your team. We often speak about personas being an outdated tool, successfully replaced by Jobs-to-Be-Done. Yet often in practice they are compatible. Both move the focus to user needs, yet they shed light onto user from different perspectives. Knowing how users think, behave and feel is as important as what they do. As Page Laubheimer noted, personas help remove box-checking mentality. They tell a story of the customer, what their environment is, what their habits are, the tools they use daily — and give product teams a way to think about users in a much more approachable and tangible way. Ultimately, use what works for you and for your team: just make sure that the user details aren’t invented, and root in actual research with actual customers. Useful resources: Personas vs. Jobs-to-Be-Done, by Page Laubheimer https://lnkd.in/eHA2Ft4J A Guide To Building Personas For UX, by Maze https://lnkd.in/ehCzACZW Personas for UX, Product, and Design Teams, by UserInterviews https://lnkd.in/eeE3pVUK A Simple Guide To Personas, by Rikke Friis Dam, Yu Siang Teo https://lnkd.in/eRA52v5m Five-Steps Framework for Building Better Personas, by Nikki Anderson, MA https://lnkd.in/eGWpqkdz Fixing User Personas, by Jordan Bowman https://lnkd.in/eDPCr63Q Personas Make Users Memorable, by Aurora Harley https://lnkd.in/eh-PYMxc A Closer Look At Personas (A Series), by Mo Goltz https://lnkd.in/eGqbr9wy https://lnkd.in/eBDsSsaR #ux #design #research