Building a Project Management Dashboard

Simon Mylius has just updated our AI Incident Tracker dashboard, which maps all (1100+) incidents in the AI Incident Database according to the MIT AI Risk Repository’s causal and domain taxonomies, and assigns each incident a harm-severity score. Using an LLM, it processes raw incident reports, providing a scalable methodology that can be applied cost-effectively across much larger datasets as numbers of reported incidents grow. The data is output in a structured dataset and a dashboard, which you can explore to identify trends and insights. For instance, you can see - distribution of incident classifications by year - distribution of incident sub-domains by year - incidents with high direct harm severity scores by year - incidents causing severe harm in more than one harm category - distribution of harm severity scores by year This update also adds new evaluation fields for each incident, including: - 5 categories of NatSec impact: Physical Security & Critical Infrastructure / Information Warfare & Intelligence Security / Sovereignty & Government Functions / Economic & Technological Security / Societal Stability & Human Rights - A Fishbone/Ishikawa diagram that presents a number of potential causes for each incident - The primary goal of the AI system involved You can read an overview of the update in the document attached, or visit our website to explore the data.

Data quality isn’t a luxury; it’s the seatbelt in your analytics car—skip it, and the crash is inevitable. Why We Actually Care (Factors) → Business decisions: Executives trust your dashboards. Don't let them down. → ML models: Garbage in, garbage out. Your model is only as good as your data. → Pipelines: One bad field breaks everything downstream. Fix it early. → Compliance: Auditors don't accept "oops." Neither does GDPR. → Cost: Bad data means reruns, fixes, and late nights. Good data saves money. The Six Dimensions (Your Quality Checklist) → Accuracy: Does it reflect reality? → Completeness: No missing pieces. → Consistency: Same story everywhere. → Timeliness: Fresh, not yesterday’s leftovers. → Validity: Fits the rules, like a puzzle piece. → Uniqueness: No duplicates—because one identity crisis is enough! How We Actually Do It (Process) → Input validation: Stop bad data at the door. Always. → Constraints & rules: If age > 150, something's wrong. → Data profiling: Know your data before you trust it. → SLAs & SLOs: Set expectations. Measure reality. → Monitoring & alerts: Catch issues before users do. → Lineage tracking: When things break, trace it back. → Triage & RCA: Fix the bug. Fix the system. Document it. The Tools That Help (Frameworks) → Great Expectations: Write tests for your data like you test code. → Deequ: Amazon's gift to data quality. Scales beautifully. → Monte Carlo: Observability for data pipelines. Sleep better. → dbt tests: Test your transformations. Trust your models. 𝗤𝘂𝗮𝗹𝗶𝘁𝘆 𝗶𝘀𝗻'𝘁 𝗮 𝗼𝗻𝗲-𝘁𝗶𝗺𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁. 𝗜𝘁'𝘀 𝗮 𝗱𝗮𝗶𝗹𝘆 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗲. Data without quality is like coffee without beans—pointless. As data engineers, we’re not just pipeline plumbers; we’re the guardians of trust. Build systems that catch issues early and keep the flavor of truth intact. 𝘉𝘶𝘪𝘭𝘵 𝘣𝘺 𝘥𝘢𝘵𝘢 𝘦𝘯𝘨𝘪𝘯𝘦𝘦𝘳𝘴, 𝘧𝘰𝘳 𝘥𝘢𝘵𝘢 𝘦𝘯𝘨𝘪𝘯𝘦𝘦𝘳𝘴. 𝘕𝘰 𝘧𝘭𝘶𝘧𝘧, 𝘫𝘶𝘴𝘵 𝘳𝘦𝘢𝘭𝘪𝘵𝘺.

(PBIX Available) Why this simple chart works better than 90% of dashboards I see? Not because it's fancy. Because every single detail has a purpose. There are 6 intentional design choices. Let me break it down. 1. The insight is IN the title "3 out of 5 goals achieved this week". Before anyone looks at a single bar, they know: - What they're looking at - The success rate (3/5) - The time period (this week) 2. You can scan this in under 3 seconds and immediately know: "Good, good, bad, good." 3. Data Labels on the Bars: Every bar shows its exact value (100, 130, 150, etc.). Why label when you have an axis? Because: - Precision matters for action (teachers need exact scores) - It eliminates squinting at the axis - It makes the chart self-contained The axis provides scale. The labels provide exactness. Both serve a purpose. 4. The vertical dotted reference line at 115 isn't decoration- it's the goal threshold. Notice three things: - It's clearly labeled ("Goal: 115") - It's positioned where the eye naturally looks (right side) - It instantly divides performance into "met goal" vs "didn't meet goal" Without that line, you'd have to mentally calculate whether 130 is good. With it? Instant understanding. 5. Minimal Color Palette: No rainbow bars. No gradient fills. Just gray bars with color-coded outcomes. Everything is gray except: - Green checkmarks (success) - Red X's (failure) - The goal line (dark gray, neutral) 6. The Layout Hierarchy: - Student selector (who am I looking at?) - Weekly summary (how did they do?) - Daily breakdown (where specifically?) Each level answers a question. That's not accident- that's intentional information architecture. The Lesson: This chart works because someone asked: - What decision does this support? (Teacher identifying struggling days) - What's the 3-second takeaway?  (3/5 goals met) - What cognitive load can I remove? (Use Icons, labels, reference line) Most dashboards fail because they show data. Great dashboards support decisions. Download the PBIX here: https://lnkd.in/gEfYdQU9 Love this? #TheVisualBreakdown series drops every other day with a new chart deconstruction. Follow + hit the bell icon so you don't miss the next one.

If your dashboard doesn’t answer these 3 questions in under 60 seconds, it’s not helping. Project managers aren’t just building reports. We’re building visibility. We’re building alignment. We’re building trust. And too often, dashboards turn into data dumps that no one actually reads. I’ve learned this the hard way: when stakeholders don’t get what they need from your dashboard, they default to side messages, follow-up meetings, or worse, silence. That's why every dashboard should focus on just three main questions: 1. What’s on track? Let them see wins at a glance. It builds confidence. Example: “Frontend 95% done, UAT still on track for Friday.” 2. What’s at risk? Call out blockers early, before they spiral. Example: “Testing delayed due to vendor handoff, patch in motion.” 3. What needs a decision? Make choices visible so momentum doesn’t stall. Example: “Scope change approval needed, will push timeline 3 days.” Dashboards are not just for project status. They’re built with stakeholders in mind, designed to match how they think, decide, and act. And when done right? They reduce status meetings. They cut back confusion. They show stakeholders exactly what they need, when they need it. Because clarity doesn’t come from more data. It comes from asking better questions. → Found this helpful? Repost ♺ and follow Jesus Romero for grounded PM frameworks that elevate clarity and trust.

“Dashboards are dead”? Only the context-free ones. Most teams start with definitions. They write a KPI dictionary, argue about formulas, then stack charts. Start with relationships. Map what drives what. Use metric maps and driver trees to sketch causality. ↳ Then define formulas. ↳ Then design screens. ↳ Then pick visuals. Here’s the 4-layer model we use: 1) Maps & Drivers: – metrics maps – driver trees 2) Definitions: – cohorts – formulas – granularity – attribution model – validation checks 3) Information Architecture – filters – page flow – drill paths – segments – comparisons 4) Visuals & UX – chart patterns – color semantics – legends & labels – responsive layout – conditional formatting Why this order? Because “what moved?” is useless without “why.” Common traps this avoids: ✕ Glossary-first thinking. Clean formulas ≠ causal logic. ✕ Chart sprawl. More graphs ≠ more clarity. ✕ Mixed levels. Result, diagnostic, actionable in one pot. If your dashboard doesn’t explain change, it’s reporting, not analytics. Build the logic first. Then display it. #dashboards

Financial reporting should be about strategic decision-making, not manual data wrangling. Yet, finance teams still spend days pulling data, reconciling numbers, and formatting reports—only to find errors at the last minute. The process is time-consuming, prone to mistakes, and slows down critical business decisions. Robotic Process Automation (RPA) with tools like UI Path is transforming financial reporting. Instead of manually extracting, cleaning, and consolidating data, automation does it for you—accurately, in real time, and without delays. Here’s how it works: ✅ Data is automatically pulled from multiple sources (ERP, CRM, spreadsheets, banks). ✅ Reconciliations happen instantly, reducing errors and improving accuracy. ✅ Reports are generated in minutes—standardized, formatted, and audit-ready. Without automation, finance teams are stuck in reactive mode, spending 80% of their time on report preparation and only 20% on analysis. The result? Slower decision-making, frustrated CFOs, and outdated insights. A company that automated its reporting process cut preparation time by 60%—freeing up finance teams to focus on forecasting, strategy, and real business impact. If your team is still manually preparing reports, you’re already behind. It’s time to automate and turn your finance team into a real-time data powerhouse. 📩 Let’s talk about how RPA can transform your financial reporting. Drop a comment or send me a message if you’re ready to make the shift! #Automation #RPA #FinanceTransformation #CFO #FinancialReporting

🔨 What is my dashboard construction process?   ❓ Whether you start with a form or a meeting, the following question is your starting point.   🤔 "What business questions are you trying to answer?"   For us, an Intake form launches an interview.   After the business question, there is always an immediate follow up: ❓ Where are you going with that?   That's important because the requestor might be asking: ◼ The wrong question. ◼ A question that's a symptom of a larger issue. ◼ A question that has already been answered. ◼ A question that we don't have data to support.   So if all that is resolved, the following questions should be addressed if the answers are not obvious, because without three Yes's, there is little value in continuing. It can be hard to be assertive with these, but it's worth the effort because it saves everyone time. ❓ What action can you take based on the answers? ❓ Are these question high enough priority that I should build a dashboard to answer them? (generally that's answered by my management). ❓ Are the questions aligned with corporate goals?   If that goes well, then it come down to Use Cases, prototyping, and gathering the appropriate the data.   The last step has been greatly simplified by my working with a separate data engineering team.   📊For the dashboards themselves, most users struggle with articulating their needs without something to look at. The sooner you show them a wireframe the better.   👀 Once I start building, I usually review my work at least once with both my manager and teammates. So by the time the customer sees something, it's had several pairs of eyes looking at it.   I usually meet with customers at least once (or twice schedules permitting) before moving work into production. In between those meetings I might have several chat/email exchanges with A/B choices.   As far as layout, I'm a big fan of the top-down approach: Goal, KPI, Trends, Actions. The rest of my design philosophy can be seen in my "Better First Dashboard" on Tableau Public (a link is on my profile).   I hope that helps! If your process is different, tell me about it in the comments.   🏗→🧠 Build to Learn!   💭🚶♀️🚶♂️ Follow for more.   #tableau #data #analytics #VizoftheRay

📌 Most Dashboards Fail Because of Bad UX Here’s the hard truth: You can have the cleanest data and the most advanced models… But if your dashboard is confusing, cluttered, or hard to navigate? Nobody will use it. BI isn’t just about data. It’s about experience. Dashboards are in fact UX products and should be treated that way. Great dashboards don’t just “show data.” They guide attention. Simplify decisions. Reduce friction. And just like any great product, they follow strong UX principles: → Clear layout → Logical flow → Minimal cognitive load → Built for the user, not the developer Let’s break down the 3 dashboard principles that make this possible 👇 1️⃣ 𝐃𝐞𝐬𝐢𝐠𝐧 𝐖𝐢𝐭𝐡 𝐭𝐡𝐞 𝐄𝐧𝐝 𝐔𝐬𝐞𝐫 𝐢𝐧 𝐌𝐢𝐧𝐝 This is where most dashboards go wrong. They’re built from a technical perspective and not a business one. Before touching a single chart, ask: → Who is this dashboard for? → What do they care about? → What action do they need to take from it? → What single question should this dashboard answer? If a dashboard tries to do everything for everyone, it ends up doing nothing for anyone. Treat your dashboard like a product. Build it around one user persona and one decision-making flow. 2️⃣ 𝐆𝐮𝐢𝐝𝐞 𝐭𝐡𝐞 𝐄𝐲𝐞 𝐰𝐢𝐭𝐡 𝐚 𝐂𝐥𝐞𝐚𝐫 𝐋𝐚𝐲𝐨𝐮𝐭 A great dashboard feels effortless to use. You don’t need to explain how to read it because it guides the user by design. Here’s how to do it: 1) Follow a natural reading pattern (top-left to bottom-right) 2) Use consistent spacing, alignment, and visual hierarchy 3) Group related charts and KPIs together 4) Avoid visual noise (limit to 5–7 key visuals per view) Think of your dashboard like a story It should unfold logically and lead the user to an insight without them having to look for it. 3️⃣ 𝐔𝐬𝐞 𝐭𝐡𝐞 𝐑𝐢𝐠𝐡𝐭 𝐕𝐢𝐬𝐮𝐚𝐥 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐉𝐨𝐛 Just because you can use a radar chart or sunburst doesn't mean you should. The best dashboards use simple, familiar visuals that communicate clearly. Here’s a cheat sheet I use: ⤷ To show progress or results → Use Scorecards or KPIs ⤷ To show trends over time → Line Charts or Area Charts ⤷ To compare parts of a whole → Pie Charts or Bar Charts ⤷ To analyze distributions → Histograms or Bell Curves ⤷ To show multivariate complexity → Heatmaps, Bubble Charts, or Pivot Tables Here what you need to remember is prioritizing clarity over creativity. Your dashboard isn’t a dribble a piece of art. It’s a decision tool. The bottom line is: Dashboards aren’t “data displays.” They’re interfaces for decision-making. And just like a product interface, design is everything. ☑ Good UX = Faster insights ☑ Good flow = Higher adoption ☑ Good visuals = Better decisions Build with purpose. Structure with clarity. Design for people. That’s how Business Intelligence becomes actual business impact. #DataStrategy #BusinessIntelligence #DataAnalytics

4 Step Process to Make Dashboards That Don't Suck 😝 I once worked with a client where for every 10 reports or data analysis we produced, 7 went unread. I am sure this sort of thing exists everywhere. So in this post, let me share my 4 step proven approach that I follow to create dashboards (or even general data analysis) that don't suck (and get noticed instead 😮) Step 1: Ask why... 🤔 Any time a client or manager asks me to perform some data analysis, I do a detailed requirement analysis. I ask them questions like: ◉ What information do you need? ◉ Why is this information helpful to you? ◉ How do you like to see it? (what graphs / interactions / detail level)? ◉ What happens if _____ is not present in the final output? ◉ Why do you care about _____ in the report? ◉ How do you plan to use this information? ◉ What is stopping you from doing this today with existing reports? ◉ Who else is going to use this and how often? Step 2: Mock-up 🎨 Once I know what we need, I then create a mock-up design. This helps me visually represent my ideas to the clients and get buy-in. These need not fancy or anything. 9 times out of 10, I make them on a paper and take a picture to the meetings. Step 3: Get the Data, Get it Right ✅ The next step is get the data and make sure it is in the right shape / format / detail. Wrong data is the second reason why data projects fail (first is wrong requirements). That means, setting up either Power Query or SQL based automation, using right source of truth, creating repeatable data pipelines and following best practice. Step 4: Create and Iterate 👨💻 Now that all my data is in place, I create the dashboards (or reports) and share with my customers. Based on their feedback and usage patterns, I revise the requirements (step 1), design (step 2), add / change data (step 3) and create next versions (step 4). This is how I ensure that my work gets noticed and used instead of lying around in a SharePoint folder somewhere. Step 5: Sharpen my saw 🧰 Of course there is a step 5 too. Instead of staying still, I take time to practice and learn new skills, new design concepts and new ways of presenting data. Here are two ways to do that: 👉 Learn by doing. If you need a step-by-step program that emphasizes on real-world examples of data analysis and dashboard reporting, check out my Excel School or Power BI courses. You will be doing a LOT of "learning by doing" in those courses. Excel School - https://lnkd.in/gktqt68p Power BI Course - https://lnkd.in/gQr5s2mJ Challenge yourself: Participate in a contest to expose yourselves to new situations and problems. I am running a Power BI dashboard contest this month and you can test your dashboarding skills in that. Check out the contest page here 👉 https://lnkd.in/gkJFJh4G That is all for now. I hope you found these tips helpful for your work. More power to you ⚡

📊 PMO Dashboard – Bringing Clarity, Control & Decision-Making in BFSI Projects POST CONTENT: In every fast-paced BFSI project environment, having a clean, data-driven PMO Dashboard is not just a reporting tool — it’s a governance enabler. Here’s what a well-structured PMO Dashboard helps achieve: 🔹 Real-Time Project Health Tracking Milestones, deliverables, risks & issues all in one place. 🔹 Accurate Reporting & Leadership Visibility Enables data-driven decisions through visual summaries. 🔹 Risk & Compliance Monitoring Allows early detection and corrective actions. 🔹 Cross-Functional Alignment Helps Business, Tech, Finance & Vendors stay on the same page. 🔹 Audit Readiness Maintains transparent documentation for BFSI governance standards. In my experience managing PMO frameworks, dashboards have improved: ✔ Reporting accuracy ✔ Stakeholder engagement ✔ On-time delivery ✔ Audit compliance If anyone is building or improving their PMO dashboards, happy to share insights and templates. #PMO #ProjectManagement #Governance #Dashboard #Reporting #BFSI #ProjectDelivery #StakeholderManagement