Manta, an IBM Company

Data pipelines are useless without lineage. It’s a bold claim. But as our founder and CEO Tomas Kratky explains, “At its heart, data lineage represents dependencies in our systems and is a key enabler for efficient change management, making every organization truly agile. “Imagine a security architect, changing the way sensitive customer data should be captured, using data lineage to instantly assess impacts of those changes on the rest of the environment. Or an analyst using data lineage to review existing sources for an executive report to decide the best way to calculate new metrics requested by the management team. “Lack of lineage only leads to reduced enterprise agility, making your pipelines harder to change and thus useless.” Follow the link to see his full commentary in last week’s “Heard on the Street” roundup from InsideBIGDATA https://ow.ly/2lwb50PXrqT

🤖 AI is advancing— and so are its problems! CNBC reports that industry leaders such as OpenAI and Google are struggling to find a solution for the phenomenon of AI “hallucinations.” 👻 What are AI hallucinations? An AI hallucination occurs when a generative AI tool sees or generates information that doesn’t exist. A good example of this is when a chatbot like ChatGPT generates false information. 👾 What causes it? While the origin is still up for debate, one culprit could be bad data. 🦾 The generative AI models used to develop chatbots are trained on existing datasets and information. This data is used to set algorithms that the AI will use to respond to users. So if bad data invades 👾 it can allow incorrect information to get rooted in the AI’s algorithm. How can you get rid of bad data? You have to be able to find it — by leveraging metadata. Metadata is data about data. It shows where your data came from, how it’s changed and how it’s continuing to evolve. Having this information gives you full visibility of your data and helps you spot the bad data so you can remove out 💥 Have questions about AI hallucinations, generative AI, or how metadata can help illuminate data accuracy? Share ‘em in the comments! https://ow.ly/uyIM50PXrqw

ChatGPT and generative AI are still top of mind across every industry. It can be a scary topic to tackle, particularly when you take into account the risks AI hallucinations pose. IT teams in service management roles need to be aware of both the value and the challenges that generative AI presents. Our Global Chief Marketing Officer, Nancy Louisnord, put together a helpful guiding article for HDI that explores a practical task-based framework: 🙋♀️ Just Me Tasks 📖 Delegated Tasks 🐴 Centaur Tasks 💻 Automated Tasks Click the link below to learn about each of these categories and how AI can make them easier. https://ow.ly/3oZi50PXrpT

Recommended reading this week from Venture Beat: Researchers have found generative AI models need human-produced data to function. 🤖AI models trained on AI-generated content (which we’re increasingly publishing across the internet) forget the human-generated data they initially learned from, performing worse over time and generating more errors. 📉Because AI-generated information overcorrects in favor of common and probable data, using AI-generated content in training materials increases the likelihood of a model repeating and intensifying human biases, including those around underrepresented groups. 🙋The good news? Avoiding model collapse hinges on access to human-generated data — either by preserving original training sets and "refreshing" or by providing new, clean, human-generated datasets. So human-created content is essential. 🧩The challenge? Knowing what’s what — including where your data came from and where it’s been. Read more at the link and let us know what you think! https://ow.ly/8cSj50PXrpA

Data governance is crucial in healthcare, ensuring accurate, high-quality data that improves clinical decision-making and patient outcomes and prevents errors. But how can providers strengthen their strategy when small changes can have a ripple effect? The prescription 💊: data lineage. 🔍 Data lineage empowers healthcare organizations to establish a robust data governance strategy, giving their teams complete control over the healthcare data pipeline. By leveraging data lineage, teams can address common data governance challenges and unlock a range of benefits, including: ✅ Better patient care & predictive analytics ✅ Enhanced regulatory compliance ✅ Increased data security and privacy ✅ Improved operational efficiency and cost savings 💡 Embracing data lineage is a game-changer for healthcare providers looking to strengthen their data governance strategy. It unlocks better patient care, regulatory compliance, data security and operational efficiency. Check out the link for more information on overcoming common challenges in healthcare data governance. https://ow.ly/T0kt50PXrp9

“Most people who work on data science, AI, and digital transformation are painfully aware that it is often culture, not technology, that stymies their efforts.” We couldn’t agree more with that opening sentence in this excellent article from the Harvard Business Review. Manta founder and CEO Tomas Kratky has even previously discussed a lack of data culture as one of the driving forces behind high turnover for chief data officers. The article linked below highlights how one company made significant strides in creating a data-driven culture by: 🏗 Building their extended data team 🚢 Getting everyone on board 📍 Pushing innovation to the fore Plenty of additional details and key takeaways can be found in each section! https://ow.ly/iH0I50PXrmX

📈Did you know that data-driven companies are 162% more likely to significantly outperform those who don’t harness the full potential of their data? 👣An important factor in that success is being able to organize and track all your data — that’s where lineage comes into play. While this sounds easy on the surface, the role of data lineage can be quite complex, so we’d like to help you take a deep dive and gain a bit more knowledge with our Ultimate Guide to Data Lineage. Download it below to get answers to the following questions: ❓What is data lineage? 🛣️What are some different approaches to data lineage? 🚨What crucial role does metadata play? 🏷️What are the different types of metadata? 🕸️Where are the sources that data lineage can be derived from? 💯How will your organization benefit from data lineage? And more! https://ow.ly/8fgr50PXrmO

📖Our recommended reading this week, from Healthcare Dive: Q1 2023 trends in the healthcare landscape include mergers and acquisitions continuing steadily.And according to consultancy group Kaufman Hall, larger hospital mergers aren’t the only trend to watch. Partnerships are expected to ramp up, “with healthcare organizations seeking complementary partnerships as they analyze merger prospects.” 📈Nonprofit healthcare systems also recorded an uptick in deal activity. ⛰️No matter which role a company plays in the world of M&As, one thing’s true: The transitions that follow include migration of enormous amounts of data. Without visibility into your data environment — and an understanding of the impact of your changes, a hard process can become nearly impossible. 🏆Bonus: Along with our recommended reading, we heartily suggest you sign up for our 🗓️July 26 webinar all about data migration! Already booked this Wednesday? Don’t fret! You can sign up now and watch the recorded session at your leisure. ⬇️Check out links to the article and the webinar⬇️ https://lnkd.in/eh_r-Dz2 https://lnkd.in/eKsWK2gh