Filtered

🎉 Congrats to Tom Jennings on his well-deserved promotion to Chief Technology Officer here at Filtered! 🚀 Tom’s leadership and vision have been key to our product successes, and we’re excited to see what he and his team deliver next. If you're curious about the exciting things he's been working on, check out our brand new interactive demo featuring cutting-edge LLMs and semantic search technology. Fill in this short form to access a free trial of our platform (for the first time ever!) and see it for yourself. ↓ ↓ ↓ https://lnkd.in/eNqT25a8

AI will either change learning in a big way (blue curve) or completely (red). We don't know. But we do know that AI requires accessible, well-organized data and information. So we all need to get our houses in order If you have that already, you're well on the way to the exponential (or at least very steep) path to fully-functioning private-network LLMs, autonomous, agents, etc. But if you don’t, you need to get there or you and your org will be left behind. I'll share links for further reading for each in the comments. 1. Quickly establish an imperfect but usable and extensible skills taxonomy. With definitions and levels. Most of these look the same. Don’t dither for months / years as so many companies do! 2. Get workable metadata for all your learning content: title, description, tags, levels, etc. 3. Make hard decisions on what to evaluate. Often, SCORM files, videos, PDFs are so poorly named you have no idea what's in them. So a) automate the opening of all files, systematically (quite hard) or b) be selective and run a manual project for a subset of the content or c) simply deprioritize any content with poor/missing data. 4. Archive the content you don't need. There's likely to be a long tail of this, especially internal content. It's clogging up the systems - computerized and human. Less is more is the well-worn cliché it is for a good reason! Humans don’t deal with overwhelming volumes of mediocre content very well. 5. Ensure ready (and ideally real-time) access to any user data e.g. activity. This will not just come from your learning systems. Analytics software like PowerBI may help here. Overlay with anecdotal, qualitative data like shares and survey results. 6. Develop hunches. If you've rolled your sleeves up and dug into the data, it's likely you'll have already developed some. It's more fruitful (and rewarding for humans!) to have AI validate hypotheses we've developed rather than leave it all to the machines. If your data isn't ready, you won't even be able to talk to IT when the next AI innovation/experiment is being discussed. You will not have a voice. Very few companies have done their L&D housekeeping. Be one of the few that has. And then, what happens next won't even be down entirely to you. But at least you'll be prepared to work with other departments (IT, Strategy, business functions) to contribute what your organization needs. Get your data in order! ♻️ Please REPOST if people you’re connected to may like this. ➕ Follow Marc Zao-Sanders for more of this kind of thing. #AI #learningdata #metadata

My latest (#15!) HBR is just out. This is my favourite. I think it’s the best-written and it touches on much of what I care about, professionally: #learning, #skills, #AI, #intention, and what it is to be human in the face of all the tech/data/cultural change. It’s about four areas in which humans still have the edge over AI (with an easy-to-remember mnemonic, ECPC): emotion, complexity, physicality, creativity. I think this is a useful framing for AI and L&D people, and certainly for those of us trying to bridge the two. Link to article in the comments. And please tell me what you think there too (HBR disabled its comments feature some time ago). ♻️ Please REPOST if you think people in your network might find this interesting. ➕ Follow Marc Zao-Sanders for more of this kind of...creativity.

i enjoyed talking about AI, personalized learning and impact with Daniel Hayward on Go1's new L&D-in-20 podcast series. if you have 20 mins, links are in the comments.

🚀 Launching end-user capabilities to tackle #ContentChaos 🚀 Workplace learning is broken. There’s more content than ever, more skills to develop, and more platforms to navigate—yet employees are still stuck in the same cycle: endless searches, irrelevant results, and no clear direction. We’ve been tackling this chaos for years with our content value analysis, AI-powered skills frameworks, advanced content tagging, precision curation, and AI powered chat agents. But today, we’re taking it to the next level: introducing end-user capabilities to our Anti-Chaos Learning Platform. The problems L&D is facing: ❌ Outdated skills frameworks: static models can’t keep up with today’s rapid skill evolution. ❌ Bloated content libraries: too much irrelevant content wastes time and budgets. ❌ Poor discoverability: employees can’t find what they need, leading to disengagement. ❌ Disconnected learners: outdated, irrelevant content fails to engage today’s workforce. The Anti-Chaos solution: 💡 Dynamic skills frameworks: skills frameworks that evolve with your business needs. 📚 Fewer, better resources: streamlined libraries focusing only on what matters. 🎯 Search that works: ultra-precise search powered by AI tagging and relevance filters. 🚀 Action-oriented end-user functionality: personalised pathways, bite-sized modules, and actionable learning resources that drive impact. This isn’t just more features—it’s the missing piece. With end-user delivery, we’re connecting L&D strategy to end-learner success and finally closing the loop on Content Chaos. Ready to rethink workplace learning? 👉 Let’s talk: https://lnkd.in/eNqT25a8 #LearningAndDevelopment #WorkplaceLearning

A year ago I posted a link to some research that highlighted four skills that were emerging as key to productivity in a post-LLM world, beyond the hype and one-off applications. Here are some resources I can recommend to help develop them.

We've collated some common "L&D pain points" that affect our customers, in case our experience in alleviating them can help you too...

The industry’s most accurate, honest, and rich search solution. Our first customer is already harnessing its power to self-curate the most relevant content, empowering admins to pinpoint hyper-relevant resources with ease. With polished metadata, relevance scores, and unparalleled accuracy, it’s search like you’ve never experienced before. Filtered’s search API doesn’t just find—it guides. See how it transforms user experiences and solves the fire-hose problem of irrelevant results. Ready to bring laser-guided precision to your platform? Let’s talk: https://lnkd.in/eNqT25a8 #contentchaos #contentdiscovery #learninganddevelopment #searchAPI

As the 2024 Learning events calendar draws to a close, we have been speaking with this years attendees and sponsors of World of Learning, Learning Technologies Exhibition & Conference, Association for Talent Development (ATD) and The LPI (Learning and Performance Institute) to name just a few. Its been an important year for these high quality (and highly attended) conferences as organisations try and match requirements centred on #skills, #automation and #AI with the latest learning tech vendor feature set and leading edge thinking. However one clear theme kept on coming up. Users and buyers are totally confused by the vendor market narrative in 2024, and particularly lost on how to tell the difference between an LMS and an LXP. How are organisations meant to make informed buying decisions when all providers appear to be saying exactly the same thing? So, if this sounds familiar - our latest research paper "The Future of the LXP" is now available and will guide you through the challenge here. This is essential reading for anyone preparing to come market (or if the process is already underway), if you're discussing how to optimise your learning provision or just feel like you are slipping behind the pace of technological change. Very interested in your thoughts Mike Zealley, Louise Scott-Worrall, Alex Ball FLPI, Alejandro Modarelli, Holly Best, Mark McCarney, Rosanna Ravey, Mark Williamson, Dion Gittoes, Eoghan Thompson, Fiona Burns, Jason Galea, David Smallbone, Adam Caldwell, Jonathan Smith KPMG UK KPMG US #kpmglearningservicesJoe Farrell Eric Thalasinos

I'm exceptionally happy to be teaming up with Dr Ashwin Mehta MBA FLPI to dig into the challenges that genAI will throw up for talent and content strategy (and just, the world, generally!) at #LNConnect24. He knows the research inside and out. I'm here to be aggravating. Topics like: - JUNK: is our codebase suitable for AI bolt ons and, more profoundly, can we break our habit of pushing junk food content on our people and calling it learning? - COST OF SKILLS: in 2025 expect CFOs to question the use of "marketing metrics" like clicks and completions in learning when they've been sold skills. Measuring the cost of skills, and therefore validating skills, will be central - LOCAL LANGUAGE MODELS. The world's greatest yet learning tool it may be, but our thesis is that lasting organisational and social benefit from genAI actually depends on tuning it with contextual and local knowledge. It's in London on Friday, hope to see some of you there.