AI For Enhancing Data Visualization

I built an AI Data Visualization AI Agent that writes its own code...🤯 And it's completely opensource. Here's what it can do: 1. Natural Language Analysis ↳ Upload any dataset ↳ Ask questions in plain English ↳ Get instant visualizations ↳ Follow up with more questions 2. Smart Viz Selection ↳ Automatically picks the right chart type ↳ Handles complex statistical plots ↳ Customizes formatting for clarity The AI agent: → Understands your question → Writes the visualization code → Creates the perfect chart → Explains what it found Choose the one that fits your needs: → Meta-Llama 3.1 405B for heavy lifting → DeepSeek V3 for deep insights → Qwen 2.5 7B for speed → Meta-Llama 3.3 70B for complex queries No more struggling with visualization libraries. No more debugging data processing code. No more switching between tools. The best part? I've included a step-by-step tutorial with 100% opensource code. Want to try it yourself? Link to the tutorial and GitHub repo in the comments. P.S. I create these tutorials and opensource them for free. Your 👍 like and ♻️ repost helps keep me going. Don't forget to follow me Shubham Saboo for daily tips and tutorials on LLMs, RAG and AI Agents.

Researchers at the Universitat Politècnica de Catalunya trained neural networks on thousands of archival sketches and photographs of Gaudí’s buildings. The models can now “guess” how an unfinished curve might wrap into a column or how a roofline could rise if the architect had kept sketching. In blind tests, some architects rated these AI completed designs as more imaginative than Gaudí’s published drawings. Apple’s recent GAUDI research project also converts a single line drawing or a short text prompt into an immersive 3D scene you can walk through in VR. The system combines two AI techniques: - Diffusion models to fill in missing visual detail, and - NeRFs (Neural Radiance Fields) to map those details onto a volumetric space. The result is a room or an entire facade that you can orbit around, inspect from new angles, and even relight. Digital artist Sofia Crespo recently projected AI generated marine patterns onto the undulating facade of Casa Batlló in Barcelona. Interactive projections and VR tours make a 19th century visionary feel contemporary to new audiences. Gaudí sketched in silence; today, algorithms can let those sketches speak - and even improvise. If you could breathe new life into any historical artwork or structure with AI, what story would you retell? #innovation #technology #future #management #startups

In the last 3 months at Ahrefs, we analyzed over 1 billion data points across 11 studies*. Here's what we learned about AI search optimization: 1. YouTube mentions are the single strongest predictor of AI visibility (correlation: 0.737) – stronger than Domain Rating, backlinks, or any traditional SEO factor. YouTube is heavily cited in AI responses, and both Google and OpenAI train on YouTube content. 2. For a given query, AI Mode and AI Overviews reach the same conclusions 86% of the time – but cite almost entirely different sources (only 13.7% citation overlap). AI Mode responses are 4x longer and mention 3x more entities. 3. Content length has essentially zero correlation with AI citations (0.04). 53% of all AI Overview citations go to pages under 1,000 words. Writing ultra-long contentisan't necessary for AI visibility. 4. Google still sends 345x more traffic than ChatGPT, Gemini, and Perplexity combined – but ChatGPT accounts for 80%+ of all AI-driven website traffic. 5. AI Overviews have a 70% chance of changing from one observation to the next, with content lasting an average of just 2.15 days. But semantic meaning stays remarkably consistent (0.95 cosine similarity). 6. "Best X" blog lists make up 43.8% of all page types cited in ChatGPT responses. 35% of those lists come from low-authority domains. 7. 79% of blog lists cited by ChatGPT were updated in 2025, and 76% of top-cited pages were refreshed within the last 30 days. Freshness matters more than ever. 8. When asked questions without valid answers, AI systems choose fabricated content with specific numbers almost every time. ChatGPT resisted best (84% accuracy), but Grok and Copilot were fully manipulated. 9. Domain Rating correlates weakly with AI visibility (just 0.266-0.326 across platforms). Number of site pages is even weaker at 0.194. 10. 67% of ChatGPT's top 1,000 citations are essentially off-limits to marketers – Wikipedia alone accounts for 29.7%, followed by homepages (23.8%) and educational content (at just 19.4%). *i'll share all the study links in a comment!

Last quarter, I worked with the MD of a heavy equipment manufacturer who believed AI would make status reports clearer and give leadership better visibility into project progress, but while the dashboards improved and the data looked sharper, the actual profit margins did not improve because delays were still being identified too late to prevent cost overruns. By the time problems appeared in reports, the financial impact had already occurred, and in 2026, with tighter compliance requirements and thinner operating buffers, that delay between issue and action is no longer affordable. What has truly changed is not reporting quality but execution speed, because AI systems can now reallocate resources, adjust schedules, and flag bottlenecks immediately instead of waiting for weekly or monthly review cycles; in plant upgrade programs and supplier transitions, I have seen problems addressed at the point of occurrence rather than after escalation. When corrective action happens closer to where the issue starts, delivery risk declines and cycle times shorten, since decisions are triggered by live data rather than by meetings or manual coordination. The main weakness I continue to see is governance, because many AI agents operate on fragmented data sources without clear ownership of decision rights, which leads teams to override outputs they do not trust and reintroduce manual controls that slow everything down, creating a false sense of stability where dashboards remain green but margin pressure builds quietly underneath. Two mistakes appear repeatedly. The first is treating AI as an advanced reporting layer, because manufacturing projects depend on operational control rather than visibility alone, and insight does not prevent delay unless the system is allowed to act within clearly defined boundaries. The second is deploying AI without defining who owns the decisions it influences, because manufacturing plants rely on accountability structures, and when escalation paths are unclear, agents can create conflicting actions that slow adoption and reduce confidence across teams. If you are beginning this journey, start by mapping a single workflow where approvals consistently delay progress, such as change requests during shutdown planning, and introduce AI only where decision rules are already stable and measurable, while avoiding areas that depend on negotiation or human judgment.  #AIInProjectManagement #AgenticAI #ExecutiveLeadership #FutureOfWork #OperationalExcellence0 #DecisionIntelligence #EnterpriseAI #ProjectGovernance #DigitalTransformation #AIForCEOs #BusinessExecution #AIStrategy

If you’re not showing up in AI search right now, the problem might not be your website. You might be invisible in the places that train the algorithms. Google, ChatGPT, Perplexity, and even DuckDuckGo are pulling data from dozens of places to decide which brands show up. Here’s how to gain real visibility in 2025: 1️⃣ Brand presence = AI visibility Most generative search tools (ChatGPT, Gemini, Perplexity) pull from: • Directories (eg. Yelp) • Trusted third-party sites (Crunchbase, BBB) • Review platforms (Trustpilot, Google Reviews) • News outlets (TechCrunch, NYTimes) • Niche forums (Reddit, Quora) If your name isn’t mentioned across the right sources, you simply won’t be part of the answer. Start with: ✅ Industry-specific directories ✅ Review platforms ✅ Reddit, Quora, niche communities ✅ Press mentions & guest posts ✅ Listicles (e.g. “Best X in [City]”) Pro Tip: Run your target keywords through ChatGPT/Perplexity/Gemini, and scroll to citations. If you're not already listed, those citation sources become your priority. 2️⃣ Create linkable assets that work Build tools and resources people actually want to link to. Examples that consistently work: • Industry statistics reports • Free calculators • Interactive tools • Original research data These assets get you mentioned across multiple key platforms simultaneously. 3️⃣ Build topical clusters Every business needs a content map that answers: • What services do you offer? • What locations do you serve? • What problems do you solve? Then build: • Dedicated commercial pages • Supporting blog content • Internal links tying it all together This is how you prove relevance and expertise - across AI and Google alike. 4️⃣ Multi-channel content Google’s increasingly prioritising video content in their search results. To capitalize on this, shoot quick clips answering questions related to your niche.  Publish them where Google heavily crawls: • Instagram • TikTok • Facebook Shorts • YouTube Shorts These are now SEO assets. You don’t need perfect production. Just relevance and consistency. 5️⃣ Schema isn’t optional anymore AI and traditional crawlers rely on structured data to understand who you are and what you do. No schema = no context = no inclusion. Cover the basics: • Organization schema • LocalBusiness schema • Product / FAQ / Article schema • Dataset schema (if applicable) Pro tip: Use ChatGPT to write your schema in seconds. Verify with Google’s Schema Validator. The businesses that win in 2025 will: ✅ Show up across all search layers ✅ Build trust signals across the web ✅ Create content built for both humans and machines

The challenge of brand visibility has taken on a new shape the age of AI search. It is not enough to rank well in search results any more. Increasingly, buyers are asking tools like ChatGPT and Google AI Overviews direct questions and trusting the answers they get, without ever clicking a link. That changes the game for marketers and communicators - when someone asks an AI tool a question in your category, does your brand show up as the answer? Our latest research with LinkedIn shows that AI platforms often do not reward the loudest brands or the most polished marketing copy. Instead, they cite content that is credible and well structured. Some of the insights that shone through in this research:  ► AI favors content that directly answers real buyer questions ► Individual expert voices account for 77% of LLM citations rather than brands  ► Structured formats like guides, comparisons, lists, and frameworks perform especially well ► Content has to stay fresh as nearly half (48%) of all citations are less than three months old The overarching lesson here is that brand discoverability is becoming less about being found in a list of links, and more about being included in the answer itself. For PR and marketing professionals charged with solving this challenge, this all means your people, your expertise, and your content need to work together to consistently answer the questions your buyers are already asking AI tools:  ► How should they choose the right vendor? ► Which features and options really matter? ► What risks and drawbacks should they watch for? ► How can they get the best value from their choice? The brands that win in AI search will be the ones that make their expertise clear and their content structured.They don’t just show up in a list of links on a search results page — they show up as the answer to an important question. We break down the full findings, including the content formats, publishing strategies, and citation patterns driving visibility across AI platforms, in our new report with LinkedIn. Download the report here: https://lnkd.in/eF2fZXu3

🚀 Agentic Development in Power BI – where the developer sets the direction, and AI helps with the heavy lifting. 💪🤖 Imagine this: 1️⃣ You describe what you need – your requirements and the rules to follow. 👨 2️⃣An AI agent turns that into a development spec – readable, reviewable, and adjustable. 🤖 👨 3️⃣You say “Go,” and the agent starts implementing the spec. 🤖 4️⃣It autonomously checks TMDL errors and runs best practice checks, fixes issues, and iterates until it gets it right. 🤖 5️⃣You open Power BI Desktop and review the result, just like you would with a teammate’s. 👨 Power BI is uniquely positioned to enable this thanks to: ✅ Open code file formats for semantic models and reports with Power BI Project file format with TMDL and PBIR. ✅ TMDL language, which brings readability, structure, and linting – making it easier for AI agents to catch and fix issues. ✅ Community-driven tools for best practice analysis that help establish the development of quality standards AI systems must adhere to. This is not just AI-assisted development. It’s AI-collaborative development. 🤝✨ ▶️Watch this end-to-end demo where I turn a requirements doc into a fully implemented Power BI semantic model - powered by AI and GitHub Copilot. Sample GitHub repo: https://lnkd.in/dFCXzv2S There are still some gaps, and detailed requirements are essential to prevent AI hallucinations. But you can already achieve great results today, and we are continuing to improve our pro-dev features for an even smoother experience. 🚀💡 #PowerBI #AgenticDevelopment

Over the past few months, I’ve been diving into Answer Engine Optimization (AEO). Also known as GEO (Generative Engine Optimization) or LLMO (Large Language Model Optimization). Whatever name we give it, the main idea is pretty simple: How can we make our content the go-to answer for AI tools like ChatGPT, Perplexity, and Google AI Overviews? This isn’t just a future trend. It’s already changing how people search, find, and engage with content. So... how do we actually show up in these AI answers? Here’s what I’m learning and experimenting with right now: ➖ Use clear, question-style H2s and H3s (“What is…?”, “How does…?”) ➖ Lead with a short, factual answer (40–60 words), then expand with details ➖ Structure content for clarity: use bullet points, tables, comparisons, and step-by-step formats ➖ Add schema markup (FAQPage, Q&A, HowTo) to improve machine readability ➖ Mirror common user questions found in “People Also Ask” and tools like AnswerThePublic ➖ Create glossary and comparison pages for key terms, tools, or use cases Keep language clear and jargon-free to ensure AI models can understand and reuse your content I'm also exploring tools like Profound, Semrush’s AI toolkit, and Ahrefs’ AI Brand Radar to track brand presence in AI answers and using tools like Hotjar | by Contentsquare to spot traffic coming from ChatGPT or Perplexity. A few resources that have been particularly valuable: Profound’s AEO Guide for Marketers (2025) CXL’s in-depth AEO playbook Steve Toth’s thoughts on LLM Optimization Amsive’s reporting on AI-driven search behavior Andreessen Horowitz’s breakdown of the new AI search UX Araks Nalbandyan's ongoing guidance from SEO experts Of course, there are still big questions around measurement, attribution, and content sustainability. AI responses are probabilistic, and what shows up today might not tomorrow. And well, zero-click experiences mean we’ll need to rethink what success looks like beyond just site visits. That said, the direction is clear. As Stefan Maritz 🎯 from CXL put it, we’re entering an era where “being the answer, not just the link in the results,” is the new currency of digital visibility. I’m curious: are you testing AEO strategies? If so, let’s compare notes.

Databricks just released the Dashboard Authoring Agent. Here is how it changes the skills required to work with data: The skill that matters now isn't knowing 𝘩𝘰𝘸 𝘵𝘰 𝘣𝘶𝘪𝘭𝘥 𝘢 𝘥𝘢𝘴𝘩𝘣𝘰𝘢𝘳𝘥.  It's knowing 𝘸𝘩𝘢𝘵 𝘲𝘶𝘦𝘴𝘵𝘪𝘰𝘯𝘴 𝘵𝘰 𝘢𝘴𝘬. Anyone can generate visualizations in seconds. Not everyone understands which metrics actually reveal what is happening in the business. The bottleneck has shifted: From technical execution → to business judgment. When an agent can: • Author datasets • Design page layouts • Create custom calculations • Detect SQL errors • Fix widget rendering issues automatically … the competitive advantage moves up the stack. I gave it one prompt and watched it plan, query, build, and organize a complete dashboard. At this point, manual widget creation becomes low-value work. What becomes high-value is understanding your data - its lineage, its gaps, and whether you can actually trust what it's telling you. Data fluency becomes the edge. If you work with data, go try it. We are genuinely redefining what's possible with data. The people who lean into that curiosity will be the ones shaping what comes next.

Hot take (fight me in the comments): Most “AI presentation tools” are just PowerPoint with confidence issues. They promise speed. Then dump a generic deck on you... And leave you fixing spacing, layouts, and structure for hours. That’s not AI. That’s outsourcing bad design. I tried Dokie AI expecting the same. It wasn’t. It built a complete investor pitch in 15 minutes. Actually used it in a real meeting. Didn't break. Here’s why it actually annoyed me (in a good way): • The slides are usable immediately • The layout doesn’t break when you edit • You can literally talk to the deck to change logos, footers, sections • It sticks to your content instead of hallucinating fluff • Slides aren't static - viewers can interact with 3D models, click through data In other words: It behaves like someone who’s built real decks before. Wild concept, I know. If your “AI tool” still needs heavy cleanup, it’s not saving time. It’s just moving the work. Agree? Disagree? Tell me which AI slide tool you think actually delivers. Try it out for yourself right here: https://lnkd.in/d6437C3u #Dokie