Brian Levenson

LLMs are learning frontend from a web full of inaccessible patterns. That is a problem. In this sharp piece, Eric W. Bailey explains how to instruct LLMs to reference the ARIA Authoring Practices Guide carefully, extracting the useful guidance while avoiding blindly copying code examples. A thoughtful reminder that accessibility still requires human judgment. https://buff.ly/BP0j2If #Accessibility #AI #WebDevelopment #ARIA #InclusiveDesign #DigitalAccessibility #A11y

30

Eric Eggert challenges the idea of blindly following so-called “best practices.” Instead, he argues for critical thinking, context, and thoughtful application when building for accessibility and the web. It’s a timely reminder: what’s “best” isn’t always universal, and may even hold us back if applied without reflection. What do you think, are “best practices” a helpful guide or a barrier to real progress? https://buff.ly/8uJQncJ #DigitalAccessibility #InclusiveDesign #UX #WebDevelopment #A11y #AccessibilityMatters #BestPractices #Accessibility #WCAG

1

Fascinating share: Cornell’s latest research shows AI acting as a group facilitator — helping more voices be heard and ideas surface (arxiv.org/abs/2508.08242). Could this shift make meetings less about outcomes and more about learning and collaboration in real time? Curious about other's experience with AI-meeting facilitation. One early example: https://lnkd.in/gQKW_xBp

4

Over my career, I’ve had the opportunity to work at the intersection of AI, data science, and K–12 education. This space is rapidly evolving, but its impact depends on how well we align innovation with real instructional needs. At Imagine Learning, we focus on building Curriculum-Informed AI, where every output is grounded in high-quality instructional materials, academic standards, and the daily realities of the classroom. Our goal is not to automate education, but to support educators with tools that save time, deepen learning, and promote equity. Whether it’s identifying a student misconception, suggesting differentiated prompts, or streamlining lesson planning, we believe AI should act as a co-pilot for teachers. It should reflect educational values, not just technical capability. I’m excited to continue shaping systems that are responsible, explainable, and instructionally sound. If you're working at the intersection of AI, education, or learning science, let’s connect and share ideas about building with purpose. #AI #DataScience #EdTech #CurriculumInformedAI #MachineLearning #ResponsibleAI #EducationTechnology #ImagineLearning #HumanCenteredAI

15

The customer is not always right, but they’re never wrong. That tension is at the heart of product work. We’re taught to listen. Talk to users. Run discovery. Prioritize outcomes. That’s the JTBD, customer-informed, value-driven approach. It’s where good product work begins. But then come the quotes from Jobs, Ford, and others. "If I had asked people, they would have asked for faster horses." "People don’t know what they want until you show it to them." Suddenly, we are told to stop listening and start leading. Invent the future. Anticipate what no one is asking for. It creates a real pull between two mindsets: 🟣 Listen first vs. lead first 🟣 Servant to the customer vs. shaper of the market 🟣 Validation vs. invention The truth is, most great product work sits in the middle. The best PMs are not order takers or solo geniuses. They are translators. They take what users say, find what they actually need, and build something that creates real progress. They listen hard but don’t treat every request as gospel. They lead with a point of view but stay grounded in real-world problems. Being customer-led doesn’t mean building everything customers ask for. Being visionary doesn’t mean ignoring the people you are here to serve. It is not either-or. It is a tension. And learning to work inside that tension is the real job. Curious how others think about this. Where do you follow feedback, and where do you trust your gut? #productmanagement #JTBD #productstrategy #leadership #innovation

15

2 Comments

Back-to-School isn’t just for students, California State University's IT team streamlines returning to campus with Smartsheet, reclaiming 30 hours per month. Read the full customer story here: https://ow.ly/FZNq30sPua4 #WorkWithFlow

4

I've partnered with many early-stage tech CEOs. The most successful ones—the true "10x" leaders—don't just have a great idea. They have a distinct set of leadership mindsets. A great article from Michael Neuendorff, PCC at Bay Area Executive Coach recently broke down 10 of these "CEO material" traits. It got me thinking about how these qualities directly amplify growth and innovation, especially when paired with a strong financial partner. As a CFO, my role isn't to be the "No" police. It's to be a strategic partner who connects the CEO’s vision to financial reality. Here’s how it works: 1. 𝗧𝗵𝗲 𝗖𝗘𝗢'𝘀 "𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗩𝗶𝘀𝗶𝗼𝗻" + 𝗧𝗵𝗲 𝗖𝗙𝗢'𝘀 "𝗙𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗥𝗼𝗮𝗱𝗺𝗮𝗽" I worked with a CEO who had a bold (and expensive!) vision for a new product line. Instead of just looking at the cost, we partnered to model the unit economics and cash-on-cash return. We built a stage-gated plan that funded the vision, proved the concept, and gave investors the confidence to double down. The CEO's vision + the CFO's financial roadmap = accelerated, de-risked innovation. 2. 𝗧𝗵𝗲 𝗖𝗘𝗢'𝘀 "𝗥𝗲𝘀𝗶𝗹𝗶𝗲𝗻𝗰𝗲" + 𝗧𝗵𝗲 𝗖𝗙𝗢'𝘀 "𝗥𝘂𝗻𝘄𝗮𝘆 𝗠𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁" Another CEO faced a brutal market shift. Their resilience was incredible, but resilience needs runway. My job was to immediately find that runway. By re-forecasting, managing burn, and communicating transparently with the board, we gave their new strategy the time it needed to work. That's the partnership: The CEO provides the grit; the CFO provides the financial breathing room. For my friends in PE, executive search, and at the helm of tech companies, this partnership is key. The best outcomes happen when a CEO's vision is amplified by a financial framework that enables, rather than restricts. That's how you build companies that last. For more insights, check out the full article here: https://lnkd.in/gjfN4PDp #CFO #CEO #Leadership #TechStartups #GrowthMindset #Innovation #PrivateEquity #ExecutiveSearch #ExecutiveTeams

6

1 Comment

The white paper, “AI + Learning Differences: Designing a Future with No Boundaries,” from Stanford Accelerator for Learning offers a thoughtful, human-centered look at how AI can expand access and equity in learning. It introduces a set of guiding principles — Privacy, Investment, Variability, Opportunity, Time, and Co-Design — with recommendations for developers, researchers, educators, and policymakers to ensure AI development is ethical, equitable, and effective for all learners. For developers, that means building with learner variability and neurodiversity in mind from the start, not as an afterthought. Today’s multimodal AI systems can process speech, text, gestures, and visuals — powerful capabilities that can open doors for students who learn and express themselves in unique ways. But technology's potential isn’t enough. It also takes intent — co-designing with learners and families, auditing for bias, and grounding every decision in principles of child development and rights. As a product leader, that philosophy resonates deeply. Still, I’ve seen how often it collides with reality — tight timelines, limited resources, competing priorities. The real question is: can we evolve from “move fast” to “move thoughtfully” when designing the next generation of learning tools? 🔗 Read the white paper: https://lnkd.in/gN9tJiKm #EdTech #LearningInnovation #AIinEducation #InclusiveDesign #ProductLeadership

9