Using Data To Improve Efficiency

𝗗𝗮𝘁𝗮 𝗴𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 𝗶𝘀 𝗼𝗻𝗲 𝗼𝗳 𝘁𝗵𝗲 𝗺𝗼𝘀𝘁 𝗺𝗶𝘀𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗼𝗼𝗱 𝘁𝗼𝗽𝗶𝗰𝘀 𝗶𝗻 𝗲𝗻𝘁𝗲𝗿𝗽𝗿𝗶𝘀𝗲. Because most people explain it from the inside out: policies, councils, standards, stewardship. But the business does not buy any of that. The business buys outcomes: → trustworthy KPIs → vendor and partner data you can actually use → faster financial close → fewer reporting escalations → smoother M&A integration → AI you can deploy without creating risk debt Most AI programs fail for boring reasons: nobody owns the data, quality is unknown, access is messy, accountability is missing. 𝗦𝗼 𝗹𝗲𝘁’𝘀 𝘀𝗶𝗺𝗽𝗹𝗶𝗳𝘆 𝗶𝘁. 𝗗𝗮𝘁𝗮 𝗴𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲 𝗶𝘀 𝗳𝗼𝘂𝗿 𝘁𝗵𝗶𝗻𝗴𝘀: → ownership → quality → access → accountability 𝗔𝗻𝗱 𝗶𝘁 𝗯𝗲𝗰𝗼𝗺𝗲𝘀 𝘃𝗲𝗿𝘆 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝘄𝗵𝗲𝗻 𝘆𝗼𝘂 𝘁𝗵𝗶𝗻𝗸 𝗶𝗻 𝟰 𝗹𝗮𝘆𝗲𝗿𝘀: 1. Data Products (what the business consumes) → a named dataset with an owner and SLA → clear definitions + metric logic → documented inputs/outputs and intended use → discoverable in a catalog → versioned so changes don’t break reporting 2. Data Management (how products stay reliable) → quality rules + monitoring (freshness, completeness, accuracy) → lineage (where it came from, where it’s used) → master/reference data alignment → metadata management (business + technical) → access controls and retention rules 3. Data Governance (who decides, who is accountable) → data ownership model (domain owners, stewards) → decision rights: who can change KPI definitions, thresholds, and sources → issue management: triage, escalation paths, resolution SLAs → policy enforcement: what’s mandatory vs optional → risk and compliance alignment (auditability, approvals) 4. Data Operating Model (how you scale across the enterprise) → domain-based setup (data mesh or not, but clear domains) → operating cadence: weekly issue review, monthly KPI governance, quarterly standards → stewardship at scale (roles, capacity, incentives) → cross-domain decision-making for shared metrics → enablement: templates, playbooks, tooling support If you want to start fast: Pick the 10 metrics that run the business. Assign an owner. Define decision rights + escalation. Then build the data products around them. ↓ 𝗜𝗳 𝘆𝗼𝘂 𝘄𝗮𝗻𝘁 𝘁𝗼 𝘀𝘁𝗮𝘆 𝗮𝗵𝗲𝗮𝗱 𝗮𝘀 𝗔𝗜 𝗿𝗲𝘀𝗵𝗮𝗽𝗲𝘀 𝘄𝗼𝗿𝗸 𝗮𝗻𝗱 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀, 𝘆𝗼𝘂 𝘄𝗶𝗹𝗹 𝗴𝗲𝘁 𝗮 𝗹𝗼𝘁 𝗼𝗳 𝘃𝗮𝗹𝘂𝗲 𝗳𝗿𝗼𝗺 𝗺𝘆 𝗳𝗿𝗲𝗲 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿: https://lnkd.in/dbf74Y9E

Today, I’ll spend 4 hours in meetings, send about 20 emails, and ping my team at least 40 times (sorry team!). Every one of those interactions is an example of "unstructured data” – data that holds the key to unlocking AI’s full potential for scaling companies. Here’s why. Some context - unstructured data is nothing new. Every company is sitting on a mountain of it. Think of all the meeting transcripts, emails, and messages that flow throughout your company every day. The information that hides in unstructured data can be incredibly valuable. It can reveal a lot about your customers’ goals and challenges, and which parts of your strategy are working (and which ones aren’t). But up to this point, it hasn’t been possible to extract insights from that information in a scalable way. So it remained beneath the surface. AI changes that. You can now ask questions of your unstructured data and generate insights that can transform your sales, marketing, and service teams – and your customers’ experience. – Marketing leaders, imagine getting a list of which features “wow” your customers on demos, so you can build your next campaign around them. – Sales leaders, picture a running list of the most common sales objections in a week, allowing you to coach your team to overcome them and close more deals. – Customer service leaders, what if you could generate a live report of which customers are showing signs of frustration? You could proactively reach out before they become a churn risk. Until recently, companies could only use a fraction of the information available to them. But soon, we'll be able to turn all their valuable information into concrete insights that customers can use every day. By unlocking unstructured data for the first time, we'll also unlock the full potential of AI to drive growth. Now, back to sending “unstructured data” (unwanted pings) to my team.

How to compare your eng team's velocity to industry benchmarks (and increase it): Step 1: Send your eng team this 4-question survey to get a baseline on key metrics: https://lnkd.in/gQGfApx4 You can use any surveying tool to do this—Google Forms, Microsoft Forms, Typeform, etc.—just make sure you can view the responses in a spreadsheet in order to calculate averages. Important: responses must be anonymous to preserve trust, and this survey is designed for people who write code as part of their job. Step 2: Calculate your how you're doing. - For Speed, Quality, and Impact, find the average value for each question’s responses. - For Effectiveness, calculate the percent of favorable responses (also called a Top 2 Box score) across all Effectiveness responses. See the example in the template above. Step 3: Track velocity improvements over time. Once you’ve got a baseline, you can start to regularly re-run this survey to track your progress. Use a quarterly cadence to begin with. Benchmarking data, both internal and external, will help contextualize your results. Remember, speed is only relative to your competition. Below are external benchmarks for the key metrics. You can also download full benchmarking data, including segments on company size, sector, and even benchmarks for mobile engineers here: https://lnkd.in/gBJzCdTg Look at 75th percentile values for comparison initially. Being a top-quartile performer is a solid goal for any development team. Step 4: Decide which area to improve first. Look at your data and using benchmarking data as a reference point, pick which metric you believe will make the biggest impact on velocity. To make this decision about what to work on to improve product velocity, drill down to the data on a team level, and also look at qualitative data from the engineers themselves. Step 5: Link efficiency improvements to core business impact metrics Instead of presenting these CI and release improvement projects as “tech debt repayment” or “workflow improvements” without clear goals and outcomes, you can directly link efficiency projects back to core business impact metrics. Ongoing research (https://lnkd.in/grHQNtSA) continues to show a correlation between developer experience and efficiency, looking at data from 40,000 developers across 800 organizations. Improving the Effectiveness score (DXI) by one point translates to saving 13 minutes per week per developer, equivalent to 10 hours annually. With this org’s 150 engineers, improving the score by one point results in about 33 hours saved per week. For so much more, don't miss the full post: https://lnkd.in/grrpfwrK

I have always believed if you nail one acquisition channel well, it is enough to scale in the first 1-2 years. And that will most often always be Meta as true incremental demand gen happens here Our entire growth in mixer grinders in the last 7-8 months ( from 0 to about 4 crs/month) have come only executing Meta ads ruthlessly The broad goal for us has been to pass on enough signals back to Meta so that Meta can continuously reach out to new relevant in-market audiences. And then have extremely strong creatives that ensures hook rate, hold rate and video views are high so that the in market audience also gets to understand the product USPs Two things that helped here which you can also implement today itself in your campaigns 1. Smart Exclusions from prospecting campaigns- brings highest incremental sales and relevant reach While custom audiences (e.g. video viewers, website visitors, past purchasers etc) are typically used for retargeting, you use them as exclusions in your prospecting campaigns. This ensures your budget isn’t wasted on people already in your funnel and forces Meta’s algorithm to seek fresh, in-market prospects. For example: Exclude recent website visitors (e.g., last 30 days) and past buyers to focus solely on new users who resemble your converters but haven’t yet engaged. This sharpens Meta’s machine learning to prioritize users showing intent without prior brand exposure, increasing the likelihood of reaching active seekers. Significantly helps in increasing new relevant reach 2. First Party Data for Algorithmic Training Wheels Instead of optimizing solely for purchases on the website ( the number of conversions on the website can be limited), pair the website conversions with the Conversions API to feed precise, first-party data back to Meta, enhancing audience modeling even with limited pixel signals due to privacy changes. This data can come from your CRM data, Amazon data etc. In our case, we get most of the first party data as buyers register their warranty for 1 year additional warranty With additional buyer data, the campaigns come out of learning way faster Try doing both of the above things. Your campaigns will scale profitably as well as bring incremental sales and reach

Using AI for quick answers is a solid start, but there’s one thing everyone is missing. And that’s the next step. The next step to unlock value with AI is connecting it to your actual work tools: Google Drive, Gmail, Calendar, and more. Granting AI access to your actual data means you can have it take direct action with less friction, so you’ll be able to request summaries of client conversations, search your communication history, or pull key insights all within ChatGPT, Claude, etc within seconds. And if you add automation platforms like Zapier or Make, you can trigger actions and sync hundreds of tasks across your entire workflow automatically. This is how you get real value from AI.

Have I mentioned we are data geeks?🤓🤓 Performance uncertainty remains one of the biggest barriers to wider uptake of #energy #efficiency technologies.💡 #Wind-assisted propulsion,💨 air-lubrication systems🫧 and other proven #retrofits can cut fuel use by double-digit percentages.📉 But real-world savings swing with weather, routing and operations. Without clarity on a retrofit’s actual contribution, neither shipowners nor charterers can forecast returns with confidence.🤷🏻♀️ And because we’ve always believed that #data📊 can give us the clearest truth, we set out to address this challenge.👊🏻 Our friends at Eastern Pacific Shipping Pte. Ltd. gave us access to the Pacific Sentinel, on which we installed a high-frequency data acquisition system as three suction #sails⛵️ were retrofitted onboard the MR tanker in March 2025. Calibrated sensors captured #power consumption, vessel speed, engine load, heading and wind conditions every 15 seconds. Over four months as the vessel traded spot around the Americas,🌎 we saw #weather and #performance at a fidelity far beyond the single daily datapoint in a noon report. Building on #ITTC and DNV methodologies, Global Centre for Maritime Decarbonisation (GCMD) and EPS implemented an “on-off’’ testing protocol,🎛️ comparing power consumption with the sails activated and deactivated under otherwise similar environmental and operational conditions to isolate the sails’ true contribution. Under the predominantly near-headwind conditions sampled, the vessel saw an average instantaneous power savings⚡️ of 7.2%, with a 95% confidence interval between 6.2% and 8.2%. Instantaneous savings ranged from +28% to –14%. These rare outliers highlight just how sensitive power savings are to wind speed and direction, and underscore the importance of tracking dynamic operational data.⚠️ Access report here:  https://lnkd.in/g_dRFtJp If we want to scale energy-efficiency retrofits, we must tackle performance uncertainty head-on. Shipowners won’t invest, and charterers won’t commit, if they can’t trust that the #savings will show up in their fuel bills.💵 We therefore developed a power savings polar heat map to predict energy and fuel savings with wind conditions. With 3rd-party verification, this will enable performance-linked financing of the retrofits.💰 This case study is but a first step in building that validation layer. And it ladders🪜 up to what we launched last week: #FEET — the world’s first blended-finance fund designed to support energy-efficiency retrofits through a pay-as-you-save repayment structure. Progress is incremental, and this marks a big step in the right direction.👊🏻 Together, we are stronger; together, we can💪🏻 Shane Balani, Zheng Yang Cheng 钟正扬, Bhushan Taskar, Goh Wan Ni, Pavlos Karagiannidis, Mirtcho Spassov, CFA, Mike Wilson, Rashim Berry, Cyril Ducau

Building Data Pipelines has levels to it: - level 0 Understand the basic flow: Extract → Transform → Load (ETL) or ELT This is the foundation. - Extract: Pull data from sources (APIs, DBs, files) - Transform: Clean, filter, join, or enrich the data - Load: Store into a warehouse or lake for analysis You’re not a data engineer until you’ve scheduled a job to pull CSVs off an SFTP server at 3AM! level 1 Master the tools: - Airflow for orchestration - dbt for transformations - Spark or PySpark for big data - Snowflake, BigQuery, Redshift for warehouses - Kafka or Kinesis for streaming Understand when to batch vs stream. Most companies think they need real-time data. They usually don’t. level 2 Handle complexity with modular design: - DAGs should be atomic, idempotent, and parameterized - Use task dependencies and sensors wisely - Break transformations into layers (staging → clean → marts) - Design for failure recovery. If a step fails, how do you re-run it? From scratch or just that part? Learn how to backfill without breaking the world. level 3 Data quality and observability: - Add tests for nulls, duplicates, and business logic - Use tools like Great Expectations, Monte Carlo, or built-in dbt tests - Track lineage so you know what downstream will break if upstream changes Know the difference between: - a late-arriving dimension - a broken SCD2 - and a pipeline silently dropping rows At this level, you understand that reliability > cleverness. level 4 Build for scale and maintainability: - Version control your pipeline configs - Use feature flags to toggle behavior in prod - Push vs pull architecture - Decouple compute and storage (e.g. Iceberg and Delta Lake) - Data mesh, data contracts, streaming joins, and CDC are words you throw around because you know how and when to use them. What else belongs in the journey to mastering data pipelines?

If I were rebuilding my GIS stack from scratch today, coming from traditional GIS, here’s exactly how I’d do it on Day 1. 1. Start with the interface layer Your hub for seeing and interacting with data. That's QGIS. This is your control center for everything. Just a powerful client that connects to everything else. 2. Replace the "GIS server" with a real data layer: PostGIS This becomes your source of truth. Store vector (and raster) data in a real database. Query it with SQL Control access and permissions, updates and transactions Run locally or in the cloud You now separate storage from visualization, which is the first big architectural shift. 3. Use modern storage formats, not proprietary files GeoParquet for vectors. Cloud Optimized GeoTIFF for rasters. Zarr for array based formats. Key difference: Leave large datasets in cloud storage. Pull only what you need, when you need it. These formats can be queried directly by analytical engines or selectively loaded into PostGIS. 4. Add an analytical/processing layer (this is where modern GIS really starts) Choose one or more: Apache Sedona SedonaDB DuckDB This layer handles: Large joins Aggregations Feature engineering Pipeline style transformations Outputs go back to PostGIS or to lightweight files for visualization. You’ve now built a small but powerful spatial data engineering pipeline that can run as frequently as you need. 5. Present your data in any way you want. Pick the tool that fits the audience: GIS-native: Felt, Atlas.co Web maps: MapLibre or Mapbox BI dashboards: Apache Superset Python: Geopandas, Rasterio, Leafmap Maps, dashboards, applications, all from the same underlying system. What’s the real difference with modern GIS? You choose the pieces, take what you need. You decide where complexity lives. Yes, you have to architect it yourself. But the tradeoff is freedom. No lock-in. No single vendor bottleneck. A system that grows with your needs, not against them. Curious who’s already moving in this direction, or planning to in 2026? 🌎 I'm Matt and I talk about modern GIS, earth observation, AI, and how geospatial is changing. 📬 Want more like this? Join 11k+ others learning from my newsletter → forrest.nyc

Walk into any insurance office and you’ll see it the “busy crisis.” People typing, calling, updating spreadsheets… but very little that actually moves the needle. 𝐈𝐭’𝐬 𝐭𝐡𝐞 𝐦𝐨𝐬𝐭 𝐝𝐚𝐧𝐠𝐞𝐫𝐨𝐮𝐬 𝐤𝐢𝐧𝐝 𝐨𝐟 𝐜𝐡𝐚𝐨𝐬: 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐯𝐞-𝐥𝐨𝐨𝐤𝐢𝐧𝐠 𝐬𝐭𝐚𝐠𝐧𝐚𝐭𝐢𝐨𝐧. Last week, a senior leader told me, “Our team is always swamped, yet our outcomes aren’t improving.” This is not a people problem. It’s a 𝐰𝐨𝐫𝐤𝐟𝐥𝐨𝐰 𝐝𝐞𝐬𝐢𝐠𝐧 𝐩𝐫𝐨𝐛𝐥𝐞𝐦. Insurance teams are drowning in activities that don’t create impact: 1. Endless follow-ups. 2. Duplicate documentation. 3. Manual reconciliations. 4. Reporting on reports. So you get a paradox: 𝐓𝐞𝐚𝐦𝐬 𝐥𝐨𝐨𝐤 𝐛𝐮𝐬𝐲… 𝐛𝐮𝐭 𝐭𝐡𝐞 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬 𝐬𝐭𝐚𝐧𝐝𝐬 𝐬𝐭𝐢𝐥𝐥. And the high-performing insurers are doing it right with: 𝟏. 𝐎𝐮𝐭𝐜𝐨𝐦𝐞-𝐛𝐚𝐬𝐞𝐝 𝐰𝐨𝐫𝐤𝐟𝐥𝐨𝐰𝐬 Every task tied to a single question: What business outcome does this drive....acquisition, retention, or claims experience? If the answer is unclear, it’s noise. 𝟐. 𝐓𝐡𝐫𝐞𝐞 𝐊𝐏𝐈𝐬 𝐭𝐡𝐚𝐭 𝐚𝐜𝐭𝐮𝐚𝐥𝐥𝐲 𝐦𝐚𝐭𝐭𝐞𝐫 𝐚) 𝐓𝐢𝐦𝐞-𝐭𝐨-𝐢𝐬𝐬𝐮𝐞 → how fast can a policy go live? 𝐛) 𝐅𝐢𝐫𝐬𝐭-𝐭𝐢𝐦𝐞-𝐫𝐢𝐠𝐡𝐭 𝐝𝐨𝐜𝐮𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 → fewer loops, fewer errors. 𝐜) 𝐂𝐥𝐚𝐢𝐦 𝐫𝐞𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧 𝐭𝐢𝐦𝐞 → the ultimate trust metric. Everything else is just ornamental. The silent crisis isn’t low productivity. It’s misaligned productivity. Fix the workflow, and suddenly the same team delivers twice the impact without working an extra minute. #InsuranceOperations #InsurTech #WorkflowAutomation #Productivity

When rigidity becomes the secret to agility: the Standardization Paradox Most leaders think standardization kills agility. Levi Strauss & Co.’s proved the opposite this week at #SAPSapphire 2026. I was impressed. Jason Gowans, CTO at Levi's, shared that the company now has more than 1,000 AI agents deployed across the company. That number is impressive. But it is not the real lesson. The real lesson is what had to change before the agents could scale. Levi’s is a global business. Different markets. Different habits. Different ways of working. And this creates a familiar leadership debate. Should we standardize, or should we let local teams stay flexible? Most companies choose a messy compromise. A bit of global control. A bit of local freedom. And a lot of invisible complexity. That model worked when humans were the glue. Humans could interpret exceptions. They could translate between systems. They could remember the undocumented rule. They could ask the person in another country how things “really work.” But AI agents change the equation. Agents cannot scale on tribal knowledge. They cannot guess every local exception. They cannot orchestrate a business where every market speaks a different operational language. This is the counterintuitive part. In the age of AI agents, standardization is not about control. It is about making the business legible. That is why standardization becomes a speed strategy. Levi’s gave a great example. Some wholesale orders used to take 2 to 5 days to process because they came through PDFs, emails, and Excel files, with long lists of sizes, colors, and styles. Now, with AI agents built on top of SAP, the same work can take 20 to 30 minutes. This is the kind of shift SAP is helping companies make: moving from fragmented work to AI-powered execution at scale: https://lnkd.in/eWajp4YB⁠ The company had to make work understandable enough for agents to execute it. That is the shift. And that is the only way to become an autonomous company. Where do you think your company is losing the most speed today: too much chaos, too many exceptions, or too little standardization? #SAPAmbassador #SAP #AI #AgenticAI #Leadership #DigitalTransformation #AutonomousEnterprise