<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Yvette Kondoh on Medium]]></title>
        <description><![CDATA[Stories by Yvette Kondoh on Medium]]></description>
        <link>https://medium.com/@ykondoh?source=rss-c9cdb4201f67------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*bOn9rWoBOQT-wNCrpVTsjA.jpeg</url>
            <title>Stories by Yvette Kondoh on Medium</title>
            <link>https://medium.com/@ykondoh?source=rss-c9cdb4201f67------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 17:40:58 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ykondoh/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Improving Customer Contributions Behavior in a Micro-Pensions Company]]></title>
            <link>https://medium.com/@ykondoh/improving-customer-contributions-behavior-in-a-micro-pensions-company-fd68956212d6?source=rss-c9cdb4201f67------2</link>
            <guid isPermaLink="false">https://medium.com/p/fd68956212d6</guid>
            <category><![CDATA[data-storytelling]]></category>
            <category><![CDATA[actionable-insights]]></category>
            <category><![CDATA[logistic-regression]]></category>
            <dc:creator><![CDATA[Yvette Kondoh]]></dc:creator>
            <pubDate>Sat, 23 Aug 2025 23:45:21 GMT</pubDate>
            <atom:updated>2025-08-26T02:58:17.669Z</atom:updated>
            <content:encoded><![CDATA[<h3>Improving Customer Contribution Behavior in a Micro-Pensions Company</h3><p>This blog post is a spin-off some slides I created and presented during the final stage of an interview process — to test my presentation skills to an audience with technical and non-technical stakeholders.</p><p>The context was based on a real-world business problem I had solved in the past but the data was simulated to create the presentation.</p><h3><strong>The Business Backstory</strong></h3><p>Imagine running a pension company for everyday workers, like farmers or construction folks (blue-collar jobs), where people contribute voluntarily whenever they can. The company has been around for 5 years, has over 17,000 customers (74% blue-collar), and allows flexible withdrawals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*g6DX9ll7aWk0pfe_.jpg" /></figure><p>But contributions are irregular, leading to “dormant” customers who stop paying in, hurting the business - irregular contributions mean unpredictable revenue, so predicting behavior can help target campaigns.</p><p>The stakeholders I reported to were the CEO, Marketing Head and Head of Operations — real people who need actionable insights.</p><h3>The Project Goal: Redefining “Active” Customers</h3><p>I redefined an “active” customer, focusing on blue-collar workers, to spot those likely to go dormant and nudge them with campaigns.</p><h3>The Methodology</h3><p>1. Frame the problem: Narrow to blue-collar customers.<br>2. Define solutions: a) Segment customers by contribution behavior b) Build a predictive model for next contribution day.<br>3. Clean and transform data: Demographics (e.g., age, gender), accounts (e.g., registration date), contributions (e.g., dates and amounts).<br>4. Implement: Build segments and model.</p><h3>The Initial Solution</h3><p>I defined customer segments based on contribution behavior and recommended that tailored campaigns be targeted at Lapsed, Contraction and Hibernating segments for stable contributions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qE7Tz89bMPCvXlN88sawwA.png" /><figcaption>Image by author</figcaption></figure><h3>The Problem With This Solution</h3><p>It only looked at two months, ignoring history. It also did not emphasize the severity of dormancy for customers at-risk.</p><h3>A Better Solution</h3><p>Built a predictive model (logistic regression) to forecast if the next contribution is in “30 days or less” or “after 30 days.”</p><p>Outcomes go beyond campaigns — like prioritizing at-risk customers using historical contributions data.</p><h3>The Data: What Went Into The Model</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z2IaU5pkeIIjW5ocxu9gHg.png" /><figcaption>Image by author</figcaption></figure><p>Sourced historical contributions data spanning 5 years for 2,658 blue-collar workers. It excluded one-time contributors and white-collar job customers.</p><h3>Defining the Target: What We Are Predicting</h3><p>The class of the target of high importance to predict correctly is late contributions (“after</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kwTC3hfI1pZft2LbrTD9nQ.png" /><figcaption>Image by author</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ul-oQ3ag2hv99kGM6mi8wQ.png" /><figcaption>Image by author</figcaption></figure><h3>The Input Features</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*itsmUz1h4WnbTVUKUoWhAw.png" /><figcaption>Image by author</figcaption></figure><h3>Evaluation Metric and Model Selection</h3><p>To ensure targeted interventions are focused on customers who are highly likely to make late payments, reducing wasted efforts and potential negative impacts on customers who would pay on time, <strong>precision </strong>is the metric to optimize.</p><p>Precision measures the proportion of customers predicted to make late payments who actually do make late payments. High precision ensures that when the model predicts a customer will make a late payment, they are very likely to actually do so.</p><p>I also settled on a <strong>logistic regression model</strong> to solve this binary classification problem.</p><h3>The Model Results: Did It Work?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zOhfm85elm2iZvSrnGYt_A.png" /></figure><p>Out of 798 customers the model was tested on, the model predicted approximately 86% (85.7%) correctly.</p><p>Out of the <strong>798</strong> customers the model was tested on, <strong>491</strong> actually contributed after 30 days.</p><p>Out of the <strong>457 </strong>customers the model predicts will contribute after 30 days, <strong>417 (91.2%) </strong>actually did contribute after 30 days. The model’s high precision for “after 30 days” (91.2%) means that when it flags a customer as likely to contribute late, you can trust that prediction about 9 times out of 10, avoiding the cost of misdirected efforts.</p><h3><strong>Top 5 predictive features for late contributions</strong></h3><p>1. Total number of days between contributions<br>2. Account age group (less than a year)<br>3. Biological age group (45–55 years)<br>4. Biological age group (55–60 years)<br>5. Whether customer has made at least one savings withdrawal</p><p>Biological age groups matter because older workers (45–60 years) might contribute differently.</p><h3>Actionable Insights: What We Learned</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oM9Xlp5ficoJL4xw8RaMag.png" /><figcaption>Image by author</figcaption></figure><p>About 3 out of 4 customers predicted to contribute after 30 days actually take over 90 days. This emphasizes the high risk and severity of true dormancy.</p><h3>Recommendations: What to Do Next</h3><ol><li>Design rewards for shorter contribution cycles to cut dormancy.</li><li>Use predicted days to prioritize dormant customers for custom campaigns.</li></ol><h3>The Impact</h3><p>Targeted campaigns from the model prediction increased contribution amount by 64% and contribution frequency by 61% compared to our baseline group. It was rewarding to see how data could re-engage people and help them invest in their future.</p><p>That’s it! I hope you enjoyed reading this blog post as much as I enjoyed putting it together.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fd68956212d6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Don’t Get Confused by a Confusion Matrix]]></title>
            <link>https://medium.com/@ykondoh/how-i-interpret-a-confusion-matrix-0cb37c1bf2c8?source=rss-c9cdb4201f67------2</link>
            <guid isPermaLink="false">https://medium.com/p/0cb37c1bf2c8</guid>
            <category><![CDATA[confusion-matrix]]></category>
            <dc:creator><![CDATA[Yvette Kondoh]]></dc:creator>
            <pubDate>Sat, 23 Aug 2025 20:37:10 GMT</pubDate>
            <atom:updated>2025-08-24T05:29:12.586Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/612/0*fsuKMFpBuesl2N5y" /></figure><p>In this post, I’ll share how I interpret a confusion matrix using a story of tackling employee turnover.</p><h3>The Challenge</h3><p>Imagine being an analyst at a company where 30% of employees are quitting, and HR needs a way to predict who’s next to avoid costly turnover. Your job? Build a model and measure it’s success. A confusion matrix is the key to seeing how well your predictive model performs.</p><h3>The Data</h3><p>Using data from 9,540 employees, I split it 67% for training and 33% for testing.</p><h3>Choosing a Metric</h3><p>Missing an employee who leaves (<em>False Negative</em>) costs more than predicting a departure that doesn’t happen (<em>False Positive</em>). Why? Replacing talent is pricier than retention efforts. So, I chose the <strong>F2 Measure</strong> (β=2.0), prioritizing <em>recall</em> to catch more leavers.</p><h3>The “Best” Model</h3><p>After exploring the data and addressing class imbalance (since fewer employees leave than stay), I tested several models for this binary classification problem (stay or leave). The winner? A <strong>logistic regression model</strong> with an average F2 score of 0.77. Not bad, but I knew we could do better.</p><h3>Initial Model: Confusion Matrix Interpretation</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dxny_UQIBh6_WxzR9wNWWg.png" /></figure><p>✅ Out of 3,149 employees in the test set, 905 actually left. At the <strong>default </strong>prediction threshold of 0.50:</p><ul><li>The model correctly identified <strong>731 leavers</strong> (80.8% recall).</li><li>But it missed <strong>174 leavers</strong> (False Negatives), wrongly predicting they’d stay.</li></ul><p>This wasn’t ideal — missing 174 employees who left could blindside HR, leading to costly turnover.</p><h3>Tweaked Model: Confusion Matrix Interpretation</h3><p>I fine-tuned the model by adjusting the prediction threshold to 0.25, optimizing for higher recall. The result? A boosted F2 score of <strong>0.82</strong> and a sharper confusion matrix:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8f1K9ltwWpWn9AF2WUHffg.png" /></figure><p>✅ Out of 3,149 employees in the test set, 905 actually left. At the <strong>optimal </strong>prediction threshold of 0.25:</p><ul><li>The model correctly identified <strong>831 leavers</strong> (91.8% recall).</li><li>And <strong>reduced </strong>missed leavers to just <strong>74</strong> (False Negatives).</li></ul><p>This tweak helped HR identify more at-risk employees, saving on recruitment costs.</p><h3>The Confusion Matrix Is Your Model’s Control Panel</h3><p>By analyzing the confusion matrix, I could see where my model was “confused” and adjust it to prioritize what matters most: catching employees at risk of leaving. Adjusting the prediction threshold (from 0.50 to 0.25) also boosted performance, catching 91.8% of leavers. This saved the company from unexpected turnover costs and empowered HR to act proactively.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0cb37c1bf2c8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Boost Your Productivity with These 3 Fun Hacks]]></title>
            <link>https://medium.com/@ykondoh/boost-your-productivity-with-these-3-fun-hacks-ddb8141f7887?source=rss-c9cdb4201f67------2</link>
            <guid isPermaLink="false">https://medium.com/p/ddb8141f7887</guid>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[rule-of-three]]></category>
            <category><![CDATA[themes]]></category>
            <category><![CDATA[habits]]></category>
            <dc:creator><![CDATA[Yvette Kondoh]]></dc:creator>
            <pubDate>Sat, 23 Aug 2025 09:02:24 GMT</pubDate>
            <atom:updated>2026-08-23T02:42:18.930Z</atom:updated>
            <content:encoded><![CDATA[<h3>Boost Your Productivity with These 3 Strategies</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5EjXmV34uYtxNvfU0wq-vg.png" /></figure><p>Have you ever wished you had more than 24 hours in a day simply because you could barely get through your to-do list and realized the day was over? <br>Well, that was how I felt when I was a college student. But over time, I discovered three game-changing strategies that turned my chaotic days into productive, satisfying ones. Following are three tips to supercharge your day.</p><h3>1. Theme Your Day Like a Party Planner</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wznYSFFzNTIBrqd7SzS2nQ.png" /></figure><p>Theming your day is like giving your each day its own vibe, like throwing a themed party. For example, my Thursdays are themed “Communication Thursday” because I want to focus on practicing and improving my public speaking skills — so I plan to read at least a chapter of a book and join a Toastmasters Meeting. On Fridays, I plan to organize my emails, folders, thoughts dump and to-do list so my Fridays are themed “Administrative Friday”. Sundays are when I plan out my clothes, meals, budget, work tasks and spiritual activities for the week so my Sundays are called “Planning Sunday”.</p><p>I really like this strategy of theming my days because it helps me get rid of decision barriers that slow down my productivity. Before using this strategy, I used to skip meals a lot but I hardly do so now because I plan my meals ahead.</p><h3>2. Master the Rule of Three</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VoYZzI8-xkmhUv8Z4wZcyQ.png" /></figure><p>The Rule of Three strategy helps me plan 3 main goals I would like to achieve in different blocks of time for different areas of my life.<br>The blocks of time could be as long as a year, short as a week and as immediate as a day.</p><p>For example, my “Rule of Three” for this week could be to: complete a work project I started at the beginning of the month, practice giving a speech in my toastmaster’s club and workout twice at the gym.</p><p>I like using the Rule of Three because it helps me prioritize and focus on what is both urgent and important at a time.</p><h3>3. Habit Bundling: Two Birds, One Stone</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oMK832GnyDB8Bm-L8hXmzA.png" /></figure><p>My final strategy for improving focus and productivity is called “Habit Bundling”. I need to exercise but I would rather listen to a sermon or a podcast instead so I pair up the two activities. Likewise, I don’t enjoy drinking water unless I am extremely thirsty so to make sure I drink water regularly, I fill my bottle with water every morning and keep it by my work desk so that I keep myself hydrated while working throughout the day.</p><h3>Wrap-Up: Make Every Day Count</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OwzJLoVWSi-UdPFR6jQGhQ.png" /></figure><p>We all get the same 24 hours, 7 days a week, and 365 days a year (plus one for leap years!). These three hacks — theming your day, using the Rule of Three, and habit bundling — can help you make the most of that time. They’ve turned my days from frantic to focused, and I’m sure they’ll do the same for you. So, pick one tip, weave it into your routine, and watch your productivity soar. Here’s to being the best, most productive version of yourself!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ddb8141f7887" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Maintain Focus and Deliver Business Value in Data Science Projects]]></title>
            <link>https://medium.com/wicds/how-to-maintain-focus-and-deliver-business-value-in-data-science-projects-426b412825d?source=rss-c9cdb4201f67------2</link>
            <guid isPermaLink="false">https://medium.com/p/426b412825d</guid>
            <category><![CDATA[blogathon]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[career-advice]]></category>
            <dc:creator><![CDATA[Yvette Kondoh]]></dc:creator>
            <pubDate>Sun, 27 Dec 2020 14:47:29 GMT</pubDate>
            <atom:updated>2025-08-23T23:53:02.782Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/612/0*ER3x_a8enyl1mq9_" /></figure><p>Starting a new data science project often comes with some excitement and the temptation to immediately dig into data sets to discover interesting insights and to apply techniques one has recently learned. Over time, the excitement quickly wanes off when faced with recurring challenges involving code bugs and data anomalies.</p><p>To reduce waning inspiration when executing a data science project, it is important to maintain the right mindset and the principle of progress over perfection to ensure sustained motivation from the onset to the completion of the project.</p><p>From my work experience, I have learned the important role structuring one’s thoughts and process flow plays in maintaining inspiration during a data science project, no matter the time commitment involved and the challenges encountered in the planning process. Structuring how to execute a data science project, whether an official work assignment or a pet project, has the advantage of keeping the data scientist focused, motivated, and prepared to show the progress made at any point in time during the project since some processes occur sequentially, in parallel, and even for most, iteratively.</p><p>In executing a data science project, I apply a <strong>20/60/20 rule</strong>. By this rule,</p><ul><li>20% of the project time is dedicated to thinking through and planning the methodology for analyses while understanding the business context, business problem, project requirements, and background of stakeholders I will be reporting to,</li><li>60% of the project time to actually do the data science work which often involves data cleaning, transformation, analyses, modeling and</li><li>the last 20% to think through the results from the analyses in order to accurately and clearly communicate the business value of key insights to the project stakeholders.</li></ul><p>Another tip for staying motivated during a data science project is to leverage techniques, tools, past project experiences, resources, user-defined functions you or a team member has created and used on past projects. By doing this, there is less time spent trying to reinvent the wheel to solve a problem that has already been solved and more time invested in discovering and translating key insights to recommendations that have a business impact.</p><p>Whereas staying motivated, focused, and inspired during a data science project is important, more crucial is the ability of a data scientist to effectively communicate relevant results of analyses done to project stakeholders and decision-makers to guide business strategy and actions. The art of crafting and telling compelling data stories using techniques such as the <a href="https://medium.com/lessons-from-mckinsey/the-pyramid-principle-f0885dd3c5c7">Pyramid Principle</a> and the <a href="https://medium.com/lessons-from-mckinsey/the-rule-of-3-c1cd82dbc96e">Rule of 3</a> help a data scientist refine and summarize the many insights generated from the data into a few main takeaways that stakeholders can act on to resolve the business problem the data science project attempted to solve.</p><p>Data scientists who invest time in planning key stages and resources of the data science process before executing, increase their chances of remaining motivated to provide valuable actionable insights that influence business decisions and strategies.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=426b412825d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/wicds/how-to-maintain-focus-and-deliver-business-value-in-data-science-projects-426b412825d">How to Maintain Focus and Deliver Business Value in Data Science Projects</a> was originally published in <a href="https://medium.com/wicds">WiCDS</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Soft Skills for Remote Data Scientists]]></title>
            <link>https://medium.com/@ykondoh/soft-skills-for-remote-data-scientists-668a3f78e6f4?source=rss-c9cdb4201f67------2</link>
            <guid isPermaLink="false">https://medium.com/p/668a3f78e6f4</guid>
            <category><![CDATA[soft-skills]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[remote-work]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[career-development]]></category>
            <dc:creator><![CDATA[Yvette Kondoh]]></dc:creator>
            <pubDate>Wed, 27 Nov 2019 10:46:32 GMT</pubDate>
            <atom:updated>2024-01-14T15:19:12.253Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XEE9LF34C0x-OFB6yKOYRg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@archigerosa?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">A R C H I G E R O S A</a> on <a href="https://unsplash.com/s/photos/remote-work?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>A remote data scientist is a data scientist who works virtually with a team or client who may or may not be in the same time zone. The data scientist may interact with an internal team or a client team via email, conferencing platforms like Zoom and instant messaging platforms like Slack. Either teams may comprise of other data scientists, data engineers, software engineers, project manager and an expert with the relevant business and domain knowledge about the business problem the data scientist is trying to solve using data.</p><p>While working as a remote data scientist with <a href="http://www.superfluid.io/">SuperFluid Labs</a>, I have acquired some useful soft skills which have contributed to my increased productivity and effectiveness as a data scientist. I have grouped these soft skills into three categories: effective communication skills, effective listening skills and organizational skills.</p><ol><li><strong>Effective Communication Skills:</strong></li></ol><p>a) <strong><em>Internal Team Level:</em></strong></p><p>i) Weekly updates and daily check-ins via email or other communication platforms adopted by the team such as Slack helps you communicate your to-do list for the week and the day. Like a scrum meeting, you can state what you accomplished the day before, what you intend on accomplishing today and any blockers you are facing. Stating the blockers you are facing enables you get the necessary assistance you require from other team members who may have faced similar blockers before and have since found solutions to them. Stating your plan for the week also helps you identify the exact resources you need to get, when and who to ask for help in order to accomplish your stated tasks.</p><p>ii) After meetings with your team or client’s team, type out a summary of what was discussed and ask your team members to review, make edits and add on to it so that you are all on the same page in understanding all that was discussed. Similarly, draw out the road map of the exact methodology for solving the problem-at-hand as discussed during internal meetings to ensure everyone is on the same page and knows their specific roles in accomplishing the project objectives.</p><p>iii) You should also show genuine concern for everyone on your team by being polite when communicating verbally, via email, Slack groups. Also learn to appreciate the contribution of your team by sending them a Kudos post on LinkedIn, for example.</p><p>b) <strong><em>Client Team Level:</em></strong></p><p>i) Cultivate the habit of asking relevant open-ended questions. This will allow the other party provide more context and information that will enable you understand a client’s business model, business problem, how the business data is generated and what the features in the data mean. This will also enable you and your team define the appropriate methodology for solving a client’s business problem.</p><p>ii) Send emails that are worded politely to your team or client, being very specific about what you are communicating or what you need clarification on. It is important to also understand the roles of the client’s team on the project such the project manager, data engineer and so on, so you address emails to the right people. Always copy the client leadership as well as your team in all email correspondence.</p><p>iii) Escalate issues and blockers such as data quality issues and data accessibility issues early to your project lead and client leadership. This will ensure that such issues are addressed in a timely manner so as not to affect the overall project timelines and deliverables.</p><p>iv) Effectively communicate the business value of your data science solution to stakeholders in a language they are most comfortable with — in terms of money. Also use appropriate visuals to communicate the key actionable insights derived from the data you have analyzed. Use the Pyramid Principle - top-down approach, to communicate logically and effectively. Start with the general, high-level points before diving into the details of what you want to communicate to stakeholders.</p><p>2. <strong>Effective Listening Skills:</strong></p><p>i) For remote engagements, you hardly get to see the other party on the other end of the line during a meeting. You only hear their voice so you must train yourself to listen to what they say as well as what they do not say. That way, you’ll be able to ask relevant follow-up questions and probe deeper till you clearly understand what is being communicated.</p><p>ii) Create an environment that allows you to be mindful, become genuinely curious and have undivided attention during meetings. It is a good idea to close all tabs in your browser that are unrelated to your meeting especially social media pages whose notifications often sound in the background while the meeting is ongoing.</p><p>iii) Try not to listen to reply. Listen to digest the information being shared, understand and then reply, asking more open-ended questions to get the other party to talk more to provide relevant information that will provide clarity in what is being communicated.</p><p>iv) One way to test if you really understood all you heard in a meeting is to write down your thoughts about what was discussed and share with the team to get their feedback.</p><p>3. <strong>Organizational Skills:</strong></p><p>i) Keep an organized calendar of your work-related meetings and activities as well as your personal career development activities. Schedule meetings with your client based on their time zone so there is no conflict in time when Daylight Saving Time kicks-off in your time zone.</p><p>ii) Keep track of your work activities to evaluate how much time you actually spend on creating valuable input on a project. You could track your hours using an excel sheet or a time tracking app. This will help you gauge your productivity level and help you price reasonable hourly rates for engagements as well as point you in the direction of the additional technical skills you should invest time in acquiring to boost your efficiency as a data scientist.</p><p>iii) Log-in to all meetings early. Always learn to give a heads-up via email or a communication platform when running late for a meeting. Try to do this before the meeting actually starts.</p><p>iv) To ensure you meet project timelines, it is a good idea to draw and keep to earlier self-imposed timelines.</p><p>v) Also learn to prioritize and accomplish tasks in parallel in a way that makes you more of an asset to your team because you are able to produce quality work in a timely manner.</p><p>vi) Develop the habit of pushing your codes to the project git repository daily, communicating with your team before you do so to avoid code conflicts.</p><p>vii) Finally, document all your analysis so that it is reproducible and there is ease of knowledge transfer with others.</p><p>I keep learning to improve on these skills everyday. I hope you are inspired to start practicing some of these soft skills to boost your productivity and make you more efficient as a professional problem solver, in general.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=668a3f78e6f4" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>