<p>To stay competitive, insurance organizations must scale the number of insights and value they produce from their data. In the second half of this <a target="_blank" href="https://www.datacamp.com/resources/webinars/future-of-data-science-in-insurance-1">recent webinar</a>, Sudaman Thoppan Mohanchandralal, Regional Chief Data and Analytics Officer at Allianz Benelux, discusses the key steps insurance organizations must take to fully realize data science&#39;s potential.</p>
<h1 id="from-data-science-garage-to-factory">From Data Science Garage to Factory</h1>
<p>In the webinar, Sudaman describes the current utilization of data science in the insurance industry as an analytics garage. Organizations currently recognize that big data is a priority, but they are not efficiently leveraging it across all decisions.</p>
<p><a target="_blank" href="https://www.ibm.com/thought-leadership/institute-business-value/report/insurance-data-personalization">IBM reports</a> that while 71% of insurers have products that rely on data in their portfolios, many do not have a cohesive data strategy to scale data science and machine learning. Sudaman likens the end-goal organizations should be striving for to an analytics factory. Here, organizations leverage data at scale across all decisions allowing for three key benefits:</p>
<ol>
<li>Massive financial potential: A lot of opportunities to scale the core business through machine learning use cases.</li>
<li>Advantages over new competitors entering the market: Making data-driven decisions is necessary to compete with new competitors with disruptive data and analytics-driven business models.</li>
<li>New possibilities to extend the core business: By making smarter decisions for the core business, resources open up to expand into new business models outside of today’s core. Some of these next-generation services are described by Capgemini <a target="_blank" href="https://worldinsurancereport.com/wp-content/uploads/sites/6/2020/05/World-Insurance-Report-2020.pdf">here</a>.</li>
</ol>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1640194652/image1_bmwal4.png" /></center><br>

<p>However, clear obstacles are currently present stopping insurers from scaling analytics factories. Here are a few of them listed in the webinar</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>
<h2 id="obstacle-1-unclear-governance">Obstacle #1: Unclear Governance</h2>
<p>The first step to reaching an analytics factory is clear governance. In <a href="https://www.datacamp.com/webinars/how-data-governance-enables-scalable-data-science">this webinar</a>, Aaren Stubberfield, Manager of Data Governance and AI at Ingredion Inc., describes how data governance minimizes business risk from data privacy laws by providing clarity about data. This clarity also allows for more confidence in decisions made based on this data. <a target="_blank" href="https://www.mckinsey.com/business-functions/mckinsey-digital/our-insights/designing-data-governance-that-delivers-value">McKinsey argues</a> that upwards of over 30% of employees&#39; time is wasted in some departments on non-value-added tasks when data quality and availability are poor due to a lack of governance. Clear governance is essential to creating a data science factory.</p>
<h2 id="obstacle-2-shortage-of-key-talent">Obstacle #2: Shortage of Key Talent</h2>
<p>The next step is developing data science talent within an organization. These skills can be broken down into three main roles:</p>
<ol>
<li>Data Scientist: Require a background in applied mathematics and technical skills in areas like data mining, statistics, machine learning, and computer science. Their main job is to build algorithms for valuable use cases.</li>
<li>Analytics Engineer: Typically have a background in data science with more of an emphasis on computer science and programming skills. Their role is to use the data scientist’s algorithms and make them scalable for production environments.</li>
<li>Analytics Translator: Have backgrounds in data science and business. Their function is to translate business problems into technical language and to translate model outputs into business solutions.</li>
</ol>
<p>Developing these skills at scale can be challenging with the <a href="https://www2.deloitte.com/us/en/insights/industry/technology/data-analytics-skills-shortage.html">data talent shortage</a>. Thus, organizations must look to upskill internally by creating a <a target="_blank" href="https://www.datacamp.com/community/blog/five-steps-to-building-a-learning-culture">learning culture</a>. These roles are all necessary for reaching a data science factory.</p>
<p>Once these roles are filled, projects with quick wins can be chosen to prove the potential of the factory. These quick wins can then be systematized and scaled to productionize the analytics process with full-stack data products. These benefits all come from building the right team and addressing the talent shortage internally.  </p>
<h2 id="obstacle-3-lack-of-frontline-involvement">Obstacle #3: Lack of Frontline Involvement</h2>
<p>The third step is involving frontline workers. Everyone in the organization must be data literate. Employees who work close to the customers need to understand the systems being deployed. They must be able to communicate why they are recommending a specific service for the customer or why services are priced the way that they are. Effective communication with customers, which is essential for building an analytics factory, comes from frontline involvement.  This is where building a data culture comes into play—which was covered by Sudaman in another <a target="_blank" href="https://www.datacamp.com/resources/webinars/webinar-building-data-culture">DataCamp webinar</a>.</p>
<h2 id="obstacle-4-management-attention">Obstacle #4: Management Attention</h2>
<p>Finally, management must be involved for the factory to work. They must work intimately with the data talent to drive organizational change. They should leverage their business knowledge and expertise to drive financial impact and value to the organization from the new technology that is developed from the data teams. Without proper attention from those making decisions, the benefits from the factory cannot be realized.</p>
<p>To learn more about how insurers can scale their analytics garages to factories, make sure to tune in to the <a target="_blank" href="https://www.datacamp.com/resources/webinars/future-of-data-science-in-insurance-1">on-demand webinar</a>.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>


https://www.datacamp.com/community/blog/moving-from-an-analytics-garage-to-a-factory

The insurance industry is rife with data and potential use cases. In a recent webinar, Allianz Benelux Regional Chief Data & Analytics Officer Sudaman T M outlined how in order to scale value from data, insurance organizations should look to move from “analytics garages”, to “analytics factories”. 

In a recent webinar, Allianz Benelux Regional Chief Data & Analytics Officer Sudaman T M outlined how in order to scale value from data, insurance organizations should look to move from “analytics garages”, to “analytics factories”. 

The Future of Data Science in Insurance: Moving from an Analytics Garage to a Factory

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1639437220/Fast_500_blog_01_ybvmld.png" /></center><br>

<p>DataCamp is proud to announce that for the third year in a row, we’ve been recognized on <a target="_blank" href="https://www2.deloitte.com/us/en/pages/technology-media-and-telecommunications/articles/fast500-winners.html">Deloitte’s Technology Fast 500™</a>, a ranking of the 500 fastest-growing technology, media, telecommunications, life science, fintech, and energy tech companies in North America.</p>
<p>“We are thrilled to be recognized for the third consecutive year on Deloitte’s Technology Fast 500,” said DataCamp CEO Jonathan Cornelissen. “We live in a different era now. Every industry is digitized and generating enormous volumes of data, but many organizations don’t have the right skills to put that data to work and outpace competitors. Just as companies had to develop computer literacy and software literacy in the past, now we must become data literate. Data literacy is driving the transformations of every industry and enables every department to deliver better results.</p>
<p>“To more effectively support our business customers and learners on their path to data literacy, we’ve expanded and evolved our platform to include robust certifications for data scientists and data analysts, career services and a job matching product to make it easier to hire data talent and to get hired, and we’ve built an in-browser collaboration tool to bridge the gap between learning data science and actually doing data science.”</p>
<h1 id="the-most-holistic-data-upskilling-platform">The most holistic data upskilling platform</h1>
<p>DataCamp offers the industry’s most holistic data upskilling experience for individuals and organizations to bridge the data skill gap, make it easier to hire and be hired, and improve collaboration and sharing of data insights.</p>
<h3 id="learn-the-most-in-demand-data-skills">Learn the most in-demand data skills</h3>
<p>DataCamp provides interactive data analytics and data science upskilling for every role, from non-coding essentials for business to data science and machine learning, all taught by experts in the field. Best of all, there’s no installation or download required—all courses and coding exercises run right in your browser, including interactive learning content in Python, R, SQL, and much more. As the leaders in data education, we’ve continued to expand our curriculum to include more content covering basic data literacy skills, BI tools like Tableau and Power BI, and data visualization, as well as more advanced topics like Plotly, Dask, and AWS Cloud Concepts.</p>
<p>This year, we also created an entirely new way for learners to apply their data skills and learn from others in the community with <a target="_blank" href="https://www.datacamp.com/data-science-competitions">DataCamp Competitions</a>. Each competition challenges learners to analyze datasets and solve real-world problems by submitting an analysis created in Workspace. Learners can see the work of their peers, upvote exceptional submissions, and earn cash prizes for winning submissions.</p>
<p>We’ve also made it even easier for learners and admins to navigate the DataCamp platform by organizing one’s learning experience and progress all in one place, and giving admins quick access to group management settings and organizational insights.</p>
<h3 id="learn-faster-with-precision-learning-plans">Learn faster with Precision Learning Plans</h3>
<p>Looking ahead to 2022, we’re very excited about the upcoming launch of Precision Learning Plans. This will be the fastest way for individuals and data-driven teams to grow their data skills.</p>
<p>Once live, companies will be able to customize the learning plan to meet their goals. Learners will take a short assessment to understand the skills they have today and the skills that they need tomorrow. Once complete, they’ll receive a Precision Learning Plan that’s exclusively targeting the training they need to take their data skills to the next level.</p>
<p>It’s effectively a way to fast-forward your learning. By receiving a personalized and adaptive learning path, you can reduce the time it takes to upskilll and achieve learning goals faster. That amounts to massive time savings for individual learners, because you’re not learning what you already know, and an even bigger time savings for companies when applied across an entire organization. If you’re interested in learning more about precision learning please reach out to your Customer Success Manager.</p>
<h3 id="certification-and-career-services">Certification and Career Services</h3>
<p>Despite the growing demand for data professionals, it’s difficult for individuals to demonstrate their skills to potential employers, and for companies to verify that candidates actually have the job-ready skills they need. Our <a target="_blank" href="https://www.datacamp.com/certification">Data Scientist and Data Analyst certifications</a> use a combination of timed assessments, a coding challenge, and a case study-style presentation evaluated by real experts to verify that a learner has both the technical and communication skills needed to be successful in their role.</p>
<p>Once a learner has been certified, we also provide them with critical tools and resources they need to land their dream jobs in data science. DataCamp’s <a target="_blank" href="https://www.datacamp.com/careerservices">Career Services</a> includes access to specialized career experts who understand the data science job market, resume reviews, and workshops on topics like interview preparation strategies.</p>
<h3 id="a-better-way-to-hire-and-get-hired-with-datacamp-talent">A better way to hire (and get hired) with DataCamp Talent</h3>
<p>We know that recruiting data scientists and other data professionals is extremely competitive and highly complex. We’re making it easier for organizations to find exceptional candidates with <a target="_blank" href="https://www.datacamp.com/hire-data-professionals">DataCamp Talent</a>. We help hiring managers find professionals with the specific skills you’re looking for to identify the right candidates faster, and we remove hiring biases to focus on a candidate’s skills above all else.</p>
<p>We want to make it easier for individuals to get matched with potential employers, too. DataCamp Talent allows individuals to easily create a polished professional profile on DataCamp to showcase their skills and data science portfolio, and get matched to companies who are interested in their experience.</p>
<h3 id="start-analyzing-and-exploring-data-instantly-no-setup-required-with-workspace">Start analyzing and exploring data instantly—no setup required—with Workspace</h3>
<p>We believe doing data science work should be easy, enjoyable, and collaborative, which is why we built <a target="_blank" href="https://www.datacamp.com/workspace">DataCamp Workspace</a>—an in-browser coding environment where you can apply your data skills to write real code, run analyses, and share your insights. Workspace lets you get started right away, whether you want to code from scratch using your own data, or start from one of our ready-to-use datasets or templates. Through our one-click publishing tool, you can easily share your analyses with others and create a professional portfolio featuring your work.</p>
<h3 id="measure-and-report-on-the-impact-of-your-organization-s-learning-program">Measure and report on the impact of your organization’s learning program</h3>
<p>DataCamp allows organizations to <a target="_blank" href="https://www.datacamp.com/groups/business/reporting">gain actionable insights into learning programs and measure ROI</a>. With <a target="_blank" href="https://www.datacamp.com/signal">DataCamp Signal skill assessments</a>, you can benchmark your team’s skills and continually reassess progress over time, and visualize skill strengths and gaps across learners with the Skill Matrix.</p>
<p>We also know that reporting on the impact of your learning programs takes time. That’s why we created Custom Reports, a one-click report builder that makes it easy for everyone to demonstrate your learning program’s ROI. With Custom Reports, you can create professional reports in one click from our pre-made templates, get detailed insights into your learning program, and schedule automated reports to make sure all the right stakeholders stay informed.</p>
<h3 id="integration-with-enterprise-systems">Integration with enterprise systems</h3>
<p>We make it simple for organizations to <a target="_blank" href="https://www.datacamp.com/groups/business/integration">plug DataCamp into their existing learning ecosystems</a>. Single Sign-On (SSO) provides learners with a seamless one-click sign-in experience using your existing SSO provider, and LMS/LXP integrations allow you to centralize learning activity data and add learning content to your learning management system.</p>
<h1 id="our-customers">Our customers</h1>
<p>We’re on a mission to help everyone thrive in a data-powered world, which is why DataCamp makes data literacy accessible to millions of people and businesses across the globe. We’ve grown our learners to 9 million and have more than 2,000 business customers, including Google, PayPal, HSBC, BNP Paribas, Barclays, Liberty Mutual Group,​ and more.</p>
<h1 id="our-datacamp-donates-partners">Our DataCamp Donates partners</h1>
<p>We launched <a target="_blank" href="https://www.datacamp.com/donates/">DataCamp Donates</a> to help organizations around the world bridge the data skill gap by providing one year of free access to DataCamp for their members. The global unemployment rate grew by 35% from 2019 to 2020, yet there’s a serious shortage of skilled tech workers to fill the future-facing jobs requiring data literacy. The reason is simple: a widespread lack of affordable, high-quality data education. We’ve already partnered with 125 organizations in more than 90 countries, and have donated more than $10 million in DataCamp scholarships so far. We’re currently accepting applications from new partner organizations through December 24, 2021. Click <a target="_blank" href="http://www.datacamp.com/donates">here</a> to apply!</p>
<h1 id="about-deloitte-s-2020-technology-fast-500-">About Deloitte’s 2020 Technology Fast 500™</h1>
<p>Now in its 27th year, the Deloitte Technology Fast 500 provides a ranking of the fastest-growing technology, media, telecommunications, life sciences, fintech, and energy tech companies—both public and private—in North America. Technology Fast 500 award winners are selected based on percentage fiscal year revenue growth from 2017 to 2020. Click <a target="_blank" href="https://www2.deloitte.com/us/en/pages/technology-media-and-telecommunications/articles/fast500-winners.html">here</a> to learn more.</p>


https://www.datacamp.com/community/blog/fast-500-2021

Deloitte recognizes DataCamp for the third consecutive year on the Technology Fast 500™.

DataCamp Named One of 500 Fastest Growing Companies in North America

<center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center><br>
<p>There are a lot of components to a data science project which require collaboration for successful deployment. With these connected parts, an agile approach is necessary for the development and deployment processes. This approach allows organizations to learn fast and iterate quickly. In <a target="_blank" href="https://www.datacamp.com/resources/webinars/manage-data-science-projects-effectively">a recent DataCamp webinar</a>, Brian Campbell, Engineering Manager, Internal Engineering at Lucid Software, discusses the best practices for successfully deploying data science projects.</p>
<h1 id="early-steps-to-allow-for-agile-development">Early Steps to Allow for Agile Development</h1>
<p>Brian explains that two key steps can be taken early in a project to allow for parallel progress on both the development and deployment of a model: working with baseline models and creating a prototype. While these approaches complicate timelines, they lead to significantly faster and better results.   </p>
<h3 id="work-with-baseline-models">Work with Baseline Models</h3>
<p>A helpful step to successfully shipping machine learning projects is creating a baseline model. A baseline model gives the same outputs as the final model. However, it is built before most of the development process and may be based on heuristics or random data. This baseline gives a target for future iterations to beat.</p>
<p>As the team develops new models, new results can be compared to the baseline and inform learning. The gap between the baseline and the desired states informs project requirements and helps teams identify the most valuable data and solutions for the problem.</p>
<h3 id="work-with-prototypes">Work with Prototypes</h3>
<p>Prototypes are models that ingest the same inputs and provide the same format of outputs that the final model will. This model can be informed by the baseline model. Once the prototype is ready, those in charge of deployment can begin work integrating the model into its intended purpose. The prototype enables parallel progress on the model; data scientists improve the model while the implementation team works on how the model will be packaged to customers or within a business process.</p>
<p>Prototypes only work when the final model behaves the same as the prototype. Thus, it is necessary to understand the available data and intended output format before creating the prototype. This requires some knowledge about the project beyond the initial problem formulation, and strong collaboration with <a target="_blank" href="https://www.datacamp.com/community/blog/3-types-of-collaborators-on-data-science-projects">problem experts</a>.</p>
<h1 id="real-world-use-case-of-baselines-and-prototypes">Real-World Use-Case of Baselines and Prototypes</h1>
<p>Brian discusses how his team at Lucid used baselines and prototypes to accelerate the development of a clustering model for a sticky note product used by product, design, and engineering teams. These teams often run design thinking sessions where a facilitator needs to manually cluster sticky notes of similar types under a given category. The system&#39;s goal was to reduce the time between brainstorming and discussion by automatically clustering ideas into categories and removing duplicates.</p>
<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1639449719/image3_cmvta2.png" /></center><br>

<p>The baseline model took in random ideas and generated random categories for them. Then, they created a prototype of this model to enable the implementation experts to begin working early in the development process. Over time, they improved the baseline model by leveraging natural language processing and machine learning to cluster similar sticky notes together.</p>
<h3 id="communication-is-important-with-these-models">Communication is important with these models</h3>
<p>The first iteration of this project ran into some problems because there was a lack of consistent collaboration between the deployment team and the data scientists working on the model. This lack of teamwork led to model performance issues, inaccurate testing, and a poor product experience. They were not able to realize the synergies of developing the model in parallel.</p>
<p>When they relaunched the project and collaborated effectively, they solved these issues and set realistic expectations. Collaboration was made possible by having a prototype model for the deployment team to work parallel with the data science team developing the model.</p>
<p>To understand the lessons learned from the project, and how to maneuver complex data science projects effectively, make sure to tune in to the <a target="_blank" href="https://www.datacamp.com/resources/webinars/manage-data-science-projects-effectively">on-demand webinar</a>.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>

https://www.datacamp.com/community/blog/how-to-ship-data-science-projects-quickly

Developing and deploying robust data science projects requires strong project management practices specific to data science. On a recent webinar, Brian Campbell discussed his best practices for managing data science projects effectively. Find out what they are here. 

On a recent webinar, Brian Campbell discussed his best practices for managing data science projects effectively. Find out what they are here. 

How to ship data science projects quickly

<center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center><br>
<p>The financial services sector sits on a treasure trove of data. It’s no wonder that the industry is rife with promising artificial intelligence and machine learning use cases. If rolled out successfully, AI can generate a value of up to <a target="_blank" href="https://www.mckinsey.com/industries/financial-services/our-insights/ai-bank-of-the-future-can-banks-meet-the-ai-challenge">$2 trillion in the banking industry</a> according to a McKinsey estimate.  In a <a target="_blank" href="https://www.datacamp.com/resources/webinars/scaling-ai-adoption-in-financial-services-1">recent webinar</a>, Shameek Kundu, former Group CDO of Standard Chartered and current Chief Strategy Officer at TruEra, outlined how to accelerate the value of AI in Financial Services.</p>
<p>Shameek explained that the value of data science and AI in financial services can be grouped into three major buckets:</p>
<ul>
<li><strong>Higher revenues</strong> through better customer experience and better decision making</li>
<li><strong>Lowering cost</strong> through more effective risk management and improved operational efficiency</li>
<li><strong>Uncovering unrealized potentials</strong> and business models.</li>
</ul>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1639438596/image1_a4pyy3.png" /></center>

<h1 id="ai-application-in-financial-services-is-broad-but-shallow">AI Application in financial services is broad but shallow</h1>
<p>Today, many AI projects in financial services remain in their infancy, unable to reach the end goal of full deployment due to a <a target="_blank" href="https://www.datacamp.com/community/blog/5-obstacles-standing-in-the-way-of-ai-adoption-in-financial-services">variety of reasons</a>. These can range from technical issues like limited data availability and quality, technological bottlenecks in deploying AI systems—to non-technical obstacles like the lack of data talent and limited trust in AI systems.</p>
<h1 id="addressing-the-lack-of-trust-in-ai-systems">Addressing the lack of trust in AI systems</h1>
<p>While complex machine learning models like <a target="_blank" href="https://www.datacamp.com/tracks/deep-learning-in-python">deep learning models</a> can provide profound solutions to previously intractable problems like <a target="_blank" href="https://www.nature.com/articles/d41586-020-03348-4">protein folding</a>, they are not great at explaining their predictions.</p>
<p>These black boxes pose societal risks if their results are implemented blindly in high-stake decisions like <a target="_blank" href="https://www.fatf-gafi.org/media/fatf/documents/reports/Opportunities-Challenges-of-New-Technologies-for-AML-CFT.pdf">anti-money laundering</a> and credit scoring. Not knowing how black-box models arrive at their predictions, end-users find it hard to trust their outputs, especially given high-profile incidents of <a target="_blank" href="https://www.datacamp.com/resources/whitepapers/data-literacy-for-responsible-ai">bias AI systems</a>.  </p>
<p>According to Shameek, three key actions can foster trust in AI.</p>
<h3 id="demystify-machine-learning-through-education">Demystify Machine Learning through Education</h3>
<p><i>General education for all</i>
Education is the long-term strategy to address the fear towards AI. As internal stakeholders, regulators, and customers gain a general understanding of how AI impacts them, they become more likely to accept AI.</p>
<p><i>Specialized education to address the skills gap</i>
The lack of technical talent prevents companies from implementing AI systems at scale. As such, companies aiming to extract value from their AI systems must be willing to hire the data talents or provide specialized training to upskill existing employees.</p>
<h3 id="set-up-internal-guardrails-for-ai-systems">Set Up Internal Guardrails for AI Systems</h3>
<p>Internal guardrails serve as the guidelines for implementing, evaluating, and monitoring AI systems.</p>
<p>These guardrails have three major components.</p>
<ul>
<li><strong>A set of guiding principles</strong> defines the boundaries of what AI can do within the company.</li>
<li><strong>Enablers for internal stakeholders</strong> facilitates the safe implementation of AI</li>
<li><strong>Standard tools and technologies</strong> evaluate the AI systems for fairness</li>
</ul>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1639438596/image3_iq6jur.png" /></center>

<h3 id="use-tools-and-technologies-to-improve-ai-quality">Use tools and technologies to improve AI quality</h3>
<p>Gaps in the AI lifecycle cause quality issues to creep into AI systems. Fortunately, there are tools to mitigate them.</p>
<p><i>Explaining black-box models</i>
The field of <a target="_blank" href="https://www.datacamp.com/community/news/explainable-ai-xai-d8tueh7c60j">explainable AI</a> remains an active field of research. It holds the promise to open up the black boxes, demystifying and building trust in the prediction-generation process.</p>
<p><i>Addressing the low data and labels quality</i>
A student might fail a test on an unfamiliar topic; a machine learning model can underperform if trained on the wrong data. AI systems can fail if the data used to train an AI system differs drastically from the data it sees in deployment. Microsoft’s python <a target="_blank" href="https://techcommunity.microsoft.com/t5/azure-ai/responsible-machine-learning-with-error-analysis/ba-p/2141774">Responsible AI Widget</a> can help identify and solve such problems.</p>
<p><i>Testing model transparency and fairness</i>
Models can be unfairly biased against a population, like a minority race or a particular gender, due to the hidden biases in the data used to train the model. Open-sourced solutions like <a target="_blank" href="https://github.com/fairlearn/fairlearn">fairlearn</a> and IBM’s <a target="_blank" href="https://github.com/Trusted-AI/AIF360">AI Fairness 360</a> can help to address these gaps.</p>
<h1 id="conclusion">Conclusion</h1>
<p>Building trust in AI within a company takes collaboration and time. Organizations need to upskill AI talent and set up safeguards for AI safety, while AI practitioners need to fully utilize tools available to them to make AI fair for all. Only when AI successfully wins the trust of its users can AI systems see widespread adoption.</p>
<p>If you are interested in the application of AI in financial services, make sure to check out Shameek’s on-demand webinar on “<a target="_blank" href="https://www.datacamp.com/resources/webinars/scaling-ai-adoption-in-financial-services">Scaling AI Adoption in Financial Services</a>”. It offers insights into the applications of AI and practical advice for accelerating its adoption.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>

https://www.datacamp.com/community/blog/building-trust-in-ai-to-accelerate-its-adoption

Building trust in AI is key towards accelerating the adoption of data science and machine learning in financial services. Learn how to accelerate the development of trusted AI within the industry and beyond.

Building trust in AI to accelerate its adoption

https://www.datacamp.com/community/blog/what-does-a-data-scientist-do

What does a data scientist do? Breaking down the responsibilities of data scientists

<center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center><br>
<p>“Storytelling is the most powerful way to put ideas into the world today,” says Robert McKee, a screenwriting guru. Today’s corporate world is rife with poorly designed slides cluttered with inappropriate charts, and data storytelling is the key to converting data to insights that drive actions and value.</p>
<p>Data stories are proven to be more memorable, persuasive, and engaging than statistics alone. But how do we tell data stories? Brent Dykes, the author of Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals, formulated a <a target="_blank" href="https://www.datacamp.com/resources/webinars/effective-data-storytelling-brent-dykes-1">valuable framework</a> for building data stories.</p>
<p>In this framework, data stories are made up of three components – narrative, data, and visuals. In this post, we will deep dive into each of these components of data stories.</p>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881977/image6_tvosqq.png" />Figure 1: Narrative, visuals and data are data stories that drive changes</center>

<h2 id="narrative-structure">Narrative Structure</h2>
<p>Most stories follow a common arc – a protagonist who faces a complication goes on a journey of resolving a difficulty before returning to their normal lives. Aristotle’s Tragedy Structure, Campbell’s Hero’s Journey, and Freytag’s Pyramid are variations of these common patterns in stories.</p>
<p>Data stories stand to benefit from the age-old narrative structure. A data story begins by setting the scene of the current situation, proceeds by providing insights that lead up to the central insight, and ends with relevant recommendations.</p>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881979/image11_do7ssc.png" />Figure 2. The narrative structure of a data story</center>

<h2 id="data">Data</h2>
<p>Analysts facing a deluge of data may be tempted to squeeze all the available data into one chart. Those who do so risk muddling their data stories with noise and confusing their audience.</p>
<p>That is why selecting the right data in a data story is crucial. Brent Dykes advised data storytellers to remove or aggregate secondary data points, as the examples below demonstrate.</p>
<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881981/image18_qtknkn.png" />Figure 3. Explicitly demonstrating that the revenue per customer is falling (right) is a better choice than plotting the total revenue and customer side-by-side (left)</center>

<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881981/image21_rgunou.png" />Figure 4. The market shares of the largest markets become apparent when the smallest markets are aggregated. </center>

<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881979/image13_mjexy3.png" />Figure 5. Often, less is more when it comes to showing multiple charts in one graph.</center>


<h2 id="visuals">Visuals</h2>
<p>Armed with the right structure and data, we are ready to use visuals to tell the data story. Here, Brent offers useful tips in crafting the perfect visual for the data story.</p>
<h4 id="1-choose-the-appropriate-visualization">1. Choose the appropriate visualization</h4>
<p>Visualizations differ in effectiveness based on the context of the data story. The key is to select the best chart based on the message.</p>
<p>For example,  bar charts are superior to pie charts in communicating differences in proportions (Figure 6A vs Figure 6B). On the other hand, pie charts are a better choice when illustrating the sum of proportions across categories (Figure 7A vs Figure 7B).</p>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881979/image14_p1lnbv.png" />Figure 6A and 6B</center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881980/image15_pbnhmh.png" />Figure 7A and Figure 7B</center>

<h4 id="2-calibrate-the-visuals-to-the-message">2. Calibrate the visuals to the message</h4>
<p>A proper data story arranges the visuals such that they meet the audience’s needs. In demonstrating the point, Dykes compared two charts that convey the same data optimized for slightly different goals.</p>
<p>Comparing Figures 7A with 7B, we observe how the alignment of horizontal axes makes it easier for the audience to make quick comparisons.</p>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881980/image19_pdgiid.png" />Figure 8A. A comparison of the total revenue from each country</center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881981/image22_fiaxrp.png" />Figure 8B. A comparison of the revenue of individual products across countries, especially for product B</center><br>


<p>On the other hand, Figure 9A and 9B are rearranged slightly differently depending on the goal of the visualization.</p><br>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881977/image5_kmirc8.png" />Figure 9A. A comparison across products for each segment</center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881978/image8_vh5it6.png" />Figure 9B. A comparison across segments for each product </center>


<h4 id="3-focusing-the-attention-of-the-audience">3. Focusing the attention of the audience</h4>
<p>The audience pays attention to what stands out in a chart. Highlighting the key points distills the signal from the noise, allowing the audience to garner insights from charts quickly.</p>
<p>As an example, the sparing use of colors (Fig 10B) and text (Fig 10C) can drastically improve a cluttered chart (Fig Figure 10A).</p><br>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881980/image20_usozqn.png" />Figure 10A</center>

<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881979/image9_vkxgdo.png" />Figure 10B</center>

<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881981/image23_itvwac.png" />Figure 10C</center><br>



<p>Another technique to focus the attention of the audience is to reveal the insights in stages using animations. In the following example (Figure 11), the audience is able to visualize the impact of a significant decline in profit at the start of Covid-19.</p>
<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881980/image16_iwjkw6.png" /></center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881979/image12_xgnpng.png" /></center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881977/image2_wyirpk.png" /></center>

<center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638881977/image1_vwaqwo.png" />Fig 11: Revealing insights in stages</center>


<h2 id="conclusion">Conclusion</h2>
<p>“Many bold, incredible insights will suffer a similar fate if they are not successfully molded into data stories,” Brent wrote. As such, the skill sets of data storytelling will continue to grow in importance as companies amass more data.</p>
<p>Whether you are new or a veteran in data storytelling, the webinar “<a target="_blank" href="https://www.datacamp.com/resources/webinars/effective-data-storytelling-brent-dykes">Effective Data Storytelling: How to Turn Insights into Action</a>” by Brent offers valuable frameworks and practical tips for data storytellers of all levels.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>


https://www.datacamp.com/community/blog/telling-effective-data-stories-with-data-narrative-and-visuals

Effective data storytelling enables data practitioners to cross the last mile of analytics, enabling them to drive action with their insights. In a recent webinar, Brent Dykes discussed how data, narrative, and visuals can drive effective data stories. Learn his best practices in this article.

Effective data storytelling enables data practitioners to cross the last mile of analytics, enabling them to drive action with their insights. In a recent webinar, Brent Dykes discussed how data, narrative, and visuals can drive effective data stories. 

Telling Effective Data Stories with Data, Narrative, and Visuals

<p>Today, high-performing insurance organizations are leveraging data science to advance their businesses in a competitive environment. In a <a target="_blank" href="https://www.datacamp.com/resources/webinars/future-of-data-science-in-insurance-1">recent webinar</a>, Sudaman Thoppan Mohanchandralal, Regional Chief Data and Analytics Officer at Allianz Benelux discussed the current state of data science in insurance and exciting use-cases of machine learning in the industry.     </p>
<h1 id="why-data-science-is-valuable-in-the-insurance-industry-today">Why Data Science is Valuable in the Insurance Industry Today</h1>
<p>Recently, the industry’s demand has shifted to more personalized services rather than those that appeal to mass markets. This <a target="_blank" href="https://www.mckinsey.com/industries/financial-services/our-insights/how-traditional-insurance-carriers-can-disrupt-through-personalized-marketing">McKinsey report</a> describes this shift Sudaman refers to. It explains how deep one-to-one relationships can be built with consumers at scale through strategic personalized marketing and service offerings.</p>
<p>These changes in needs and relationships are made possible by the massive growth of available data for individual customers as well as advances in machine learning and resources. In <a target="_blank" href="https://www.ibm.com/thought-leadership/institute-business-value/report/insurance-data-personalization">this article</a>, IBM discusses how insurers observe a 22-25% improvement in time to profitability when creating personalized services relative to mass-appealing products. These products can only be developed at scale using machine learning.  </p>
<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638880876/image3_lya3kn.png" /></center><br>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>
<h1 id="the-state-of-data-science-in-insurance">The state of data science in insurance</h1>
<p>Sudaman explains in the webinar that machine learning provides value when it is leveraged for predictions rather than causal inference and when the problem is sufficiently self-contained. As of today, machine learning fails when organizations attempt to use these algorithms as a crystal ball or black box rather than a tool to inform humans on directions to explore. Here’s a glimpse of some of the use-cases covered in the webinar:</p>
<h3 id="1-fraud-detection-in-insurance-claims">1. Fraud Detection in insurance Claims</h3>
<p>Machine learning performs very well on prediction tasks with a lot of data. Therefore, <a target="_blank" href="https://www.datacamp.com/courses/fraud-detection-in-python">fraud detection</a> is a great use case to train and deploy a classification algorithm such as logistic regression or a decision tree. This system can be used to flag claims that look suspicious. This makes fraud prevention significantly more manageable for an analyst as it will augment their workflow by surfacing the claims most likely to be fraudulent.</p>
<h3 id="2-identifying-customers-for-retention-activities">2. Identifying Customers for Retention Activities</h3>
<p>Another important use case for improving the scale of any organization is improving <a target="_blank" href="https://www.datacamp.com/courses/machine-learning-for-marketing-in-python">customer retention</a>. Machine learning can help to identify which customers are at risk of leaving your organization. Identifying these customers at scale and providing incentives for them to stay will lead to better customer lifetime value in the long run—enabling deeper personalization.</p>
<h3 id="3-optimizing-pricing-with-time-series-data-risk-premium-modeling-">3. Optimizing Pricing with Time Series Data (Risk Premium Modeling)</h3>
<p>Understanding a customer’s risk to file a claim is critical in determining the correct price and structure of a plan to provide them. All insurers are doing this at some level. With the expansion of available data, model accuracy to determine a customer’s risk profile can be improved to offer more <a target="_blank" href="https://www.datacamp.com/courses/life-insurance-products-valuation-in-r">competitive pricing</a> to individuals.</p>
<h3 id="4-future-disease-prediction">4. Future Disease Prediction</h3>
<p>Similar to the previous use case, understanding future disease risks is critical in portfolio optimization and pricing. <a target="_blank" href="https://www.lloyds.com/~/media/files/news-and-insight/coronavirus-hub/product-simplification-report/lloyds_product_simplification_report_final.pdf?la=en">This Lloyd’s of London report</a> discusses how the insurance industry will have to pay out upwards of $100 billion in COVID-related claims. While predicting the pandemic in advance would have been challenging, understanding and predicting health outcomes will enable insurers to prioritize and personalize their services.</p>
<h3 id="5-portfolio-optimization">5. Portfolio Optimization</h3>
<p>The final use case we will discuss is portfolio optimization. <a href="https://www.datacamp.com/courses/introduction-to-portfolio-risk-management-in-python">Portfolio management</a> consists of a few steps: identifying meaningful groupings of risk to be analyzed, informing optimal decisions, and identifying opportunities. Each of these subtasks is solvable through data science optimization techniques.</p>
<p>If you want to learn more about the future of data science in insurance, make sure to tune in to the webinar <a target="_blank" href="https://www.datacamp.com/resources/webinars/future-of-data-science-in-insurance-1">here</a>.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>

https://www.datacamp.com/community/blog/data-science-in-insurance-today

The insurance industry is rife with data and data science use-cases that provide value. In a recent webinar, Allianz Benelux Regional Chief Data & Analytics Officer Sudaman T M explored the state of data science in the industry today.

Data Science in Insurance Today

<p>Successful data science projects require collaboration across teams. Data science projects are not impactful without proper alignment with all relevant stakeholders. In <a target="_blank" href="https://www.datacamp.com/resources/webinars/manage-data-science-projects-effectively">a recent DataCamp webinar</a>, Brian Campbell, Engineering Manager, Internal Engineering at Lucid Software, discusses the best practices for cross-team collaboration on data science projects.</p>
<h2 id="three-types-of-collaborators">Three Types of Collaborators</h2>
<p>Data scientists must collaborate with other stakeholders to drive impact with their projects. Data scientists should rely on others within their organization to better understand the business value and metrics underlying the problems they are interested in, how to navigate complex data environments, and how to deploy their models to customers. In the webinar, Brian discusses three types of collaborators:</p>
<h3 id="problem-expert">Problem Expert</h3>
<p>The problem expert is the collaborator who sets the foundation for a data science project. Their job is to provide domain expertise for a project and to define the problem statement.</p>
<p>Problem experts intimately understand the metrics impacted by a data science project and how they translate to business outcomes. They also know the accuracy needed for a model to be valuable. This knowledge is critical information because no model can be perfectly accurate. Understanding what is reasonable and valuable to aim for is essential to defining goals and setting expectations early in a project.</p>
<h3 id="data-expert">Data Expert</h3>
<p>Data is the key ingredient for any successful data science project. Many organizations have complex data landscapes that make it challenging to know what data is available and what data is needed to solve a problem.</p>
<p>Data experts are in charge of knowing what data is available in a domain, how to get the data, and what data is best suited for specific business questions. They can sometimes be the problem expert themselves but are typically in data engineering roles within the organization. </p>
<h3 id="implementation-experts">Implementation Experts</h3>
<p>Finally, implementation experts are in charge of understanding how to deploy these models to customers. They are experts in the infrastructure and resources required to deploy models at scale.</p>
<p>When collaborating with these experts, it&#39;s important to let them know well in advance about the project so that they have time to ensure there is proper infrastructure in place when the model is ready to be deployed. It’s also important to know how the work will be used—as a feature in the product, part of an API, integrated with an existing process, or requiring a new process from scratch. The project&#39;s use will determine who the best implementation expert to collaborate with will be.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>
<h2 id="how-to-effectively-work-with-collaborators">How to Effectively Work With Collaborators</h2>
<p>Timelines are a critical component of collaboration. These allow everyone involved in the project to plan when they will get involved to help advance it to completion.</p>
<p>Consider a proposed project to deploy a customer retention model. This model will identify customers who are at risk of no longer purchasing a subscription from the organization and will send a promotional email to them to entice them to stay.</p>
<p>This project will require initially understanding customer retention metrics and trends, collecting relevant data, cleaning and analyzing the data, building the classification model, integrating it into the automated email system, and setting it live. All of these steps could take weeks and many steps will require help from the three stakeholders discussed in the previous section.</p>
<br><center><img src="https://res.cloudinary.com/dyd911kmh/image/upload/v1638469335/image1_teqlfr.png" /></center><br>

<p>Setting an initial timeline allows collaborators to roadmap when they need to be available to help with the project. If delays occur, partners should be informed immediately. Implementation experts should not show up a month early if there is a delay in the data collection phase as this will waste their time and hurt opportunities for future collaborations with them.</p>
<p>To learn more about data science project management best practices, tune in to the <a target="_blank" href="https://www.datacamp.com/resources/webinars/manage-data-science-projects-effectively-1">on-demand webinar</a> for more in-depth examples.</p>
<br><center><a target="_blank" href="https://www.datacamp.com/groups/business"><img src="https://res.cloudinary.com/dyd911kmh/image/upload/f_auto,q_auto:best/v1624651228/schedule_a_demo_bfhlx1.png" /></a></center>

https://www.datacamp.com/community/blog/3-types-of-collaborators-on-data-science-projects

Successfully shipping data science projects requires collaborating with various types of stakeholders. Who are these stakeholders? On a recent webinar, Brian Campbell unpacked 3 types of collaborators you can find in every data science project.

community

datacamp

Official Blog