<?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 Dhan Moti on Medium]]></title>
        <description><![CDATA[Stories by Dhan Moti on Medium]]></description>
        <link>https://medium.com/@dhanmoti?source=rss-34d60b1d76d1------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*mlwh-TsEw-65h3_j0qFy8A@2x.jpeg</url>
            <title>Stories by Dhan Moti on Medium</title>
            <link>https://medium.com/@dhanmoti?source=rss-34d60b1d76d1------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 20:09:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dhanmoti/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[Why Budgeting Alone Often Fails Us]]></title>
            <link>https://medium.com/tulsi-tech/why-budgeting-alone-often-fails-us-e6c0b9b76c61?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/e6c0b9b76c61</guid>
            <category><![CDATA[habits]]></category>
            <category><![CDATA[mindfulness]]></category>
            <category><![CDATA[personal-finance]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[self-improvement]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Fri, 23 Jan 2026 08:59:32 GMT</pubDate>
            <atom:updated>2026-01-23T08:59:32.611Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*goF5b_NnHW4QAMG9PryDwg.png" /></figure><p>Most of us don’t think of ourselves as careless with money.</p><p>We download budgeting apps.<br> We make spreadsheets.<br> We tell ourselves, <em>This month will be different.</em></p><p>And for a few days, it is.</p><p>Then life happens.</p><p>A stressful meeting.<br> A long commute.<br> A quiet evening when willpower feels expensive.</p><p>And somehow, we’re back at the same question — <br><strong>“Why do I keep overspending when I <em>know</em> better?”</strong></p><h3>The Discipline Myth</h3><p>The popular answer is discipline.</p><p>We assume that if we just tried harder — tracked more closely, restricted more tightly, planned more aggressively — we’d finally “fix” our finances.</p><p>But here’s the uncomfortable truth:<br><strong>Most money problems aren’t discipline problems.</strong></p><p>They’re awareness problems.</p><p>If discipline were the issue, budgets would work far more often than they do.</p><h3>The Moment Budgets Don’t See</h3><p>Think about your last unplanned purchase.</p><p>Maybe it was small:<br> a coffee, a snack, a ride you didn’t strictly need, a late-night online buy.</p><p>You probably didn’t sit down, review your budget, and <em>decide</em> to overspend.</p><p>It happened quickly. Almost automatically.</p><p>Later, you noticed the damage and asked: <em>“Why did I do that?”</em></p><p>That question is more important than it seems.</p><p>Because that purchase wasn’t random — it was a response.</p><p>To fatigue.<br> To stress.<br> To boredom.<br> To habit.</p><p>Budgets are designed for the future.<br> Purchases happen in the present.</p><p>And that gap is where most systems fail.</p><h3>Money Is Emotional Before It’s Logical</h3><p>We like to believe we are rational with money.</p><p>In reality, money is one of the most emotional parts of modern life.</p><p>We spend to feel comfort.<br> We spend to feel relief.<br> We spend to feel rewarded, distracted, soothed, or in control.</p><p>No spreadsheet captures that moment when you think, <em>“I deserve this.”</em><br> No category explains, <em>“I just need something easy right now.”</em></p><p>So when budgets fail, we blame ourselves — <br> instead of noticing what the behavior is trying to tell us.</p><h3>Awareness Changes the Game</h3><p>Here’s the counter-intuitive insight:</p><p><strong>Behavior changes faster through awareness than through restriction.</strong></p><p>Noticing an impulse — without judging it — creates space.<br> And in that space, choice appears.</p><p>Awareness sounds passive, but it’s not.</p><p>It’s the difference between:</p><ul><li><em>“I shouldn’t spend this” </em>and</li><li><em>“Interesting… I’m reaching for my wallet because I’m exhausted.”</em></li></ul><p>The second statement doesn’t fight reality.<br> It understands it.</p><p>And understanding dissolves patterns that force never could.</p><h3>A Small Experiment (That Actually Works)</h3><p>Try this for just three days.</p><p>No new budget.<br> No cutting back.<br> No rules.</p><p>Only this:</p><ol><li><strong>Notice every purchase.</strong></li><li><strong>Pause for a moment.</strong></li><li><strong>Ask yourself what you’re feeling.</strong></li><li><strong>Don’t judge the answer.</strong></li></ol><p>That’s it.</p><p>Some people like using a quiet, offline-first tool like <em>Tulsi</em> to log these moments, but the tool isn’t the transformation.</p><p>The noticing is.</p><p>What surprises most people isn’t guilt — it’s clarity.</p><p>They begin to see patterns:</p><ul><li>Spending spikes when they’re tired, not careless.</li><li>Small purchases cluster around stressful days.</li><li>“Impulse” buys follow emotional triggers, not poor math.</li></ul><p>And once seen clearly, these patterns start loosening on their own.</p><h3>Why This Feels Different</h3><p>Traditional budgeting says:<br> <em>“Control yourself.”</em></p><p>Awareness-based finance says:<br> <em>“Understand yourself.”</em></p><p>One creates tension.<br> The other creates insight.</p><p>And insight changes behavior quietly, sustainably, and without burnout.</p><p>You don’t stop overspending because you’re stricter.<br> You stop because the action no longer feels unconscious.</p><h3>The Real Shift</h3><p>Financial progress doesn’t begin with numbers.</p><p>It begins with attention.</p><p>When you see clearly:</p><ul><li>You spend less without forcing it.</li><li>You trust yourself more.</li><li>Money stops feeling like a constant failure loop.</li></ul><p>Awareness is free.<br> It doesn’t require motivation.<br> And it works even on bad days.</p><p>That’s why it changes everything.</p><h3>Reflection Prompt</h3><p>Take one purchase today.</p><p>Pause and ask:</p><ul><li>What was I feeling right before this?</li><li>What did I actually need in that moment?</li></ul><p>Don’t fix it.<br>Just notice.</p><p>That’s where change starts.</p><p><a href="https://tulsifinance.com/">Tulsi | Private Money &amp; Carbon Tracker</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e6c0b9b76c61" width="1" height="1" alt=""><hr><p><a href="https://medium.com/tulsi-tech/why-budgeting-alone-often-fails-us-e6c0b9b76c61">Why Budgeting Alone Often Fails Us</a> was originally published in <a href="https://medium.com/tulsi-tech">Tulsi Tech</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Optimize Corporate Credit Decisions — A Simple Keras Project]]></title>
            <link>https://medium.com/@dhanmoti/optimize-corporate-credit-decisions-a-simple-keras-project-895ae365bf0d?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/895ae365bf0d</guid>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Fri, 18 Apr 2025 13:38:57 GMT</pubDate>
            <atom:updated>2025-04-18T13:48:07.103Z</atom:updated>
            <content:encoded><![CDATA[<h3><em>Optimize Corporate Credit Decisions — A Simple Keras Project</em></h3><p>AI/ML #4: Credit Line Optimization using Deep Learning with Keras Library</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ljhaQGrnRlllO8Un3x00_w.png" /></figure><h3>Summary</h3><p>This project shows how deep learning can help banks make better credit decisions for corporate clients. Using the Keras library, I built a simple neural network model that predicts the best credit line a business should get based on its financial data.</p><p>The project uses Python, Keras, and tools like SHAP for explainability. The goal isn’t to replace humans but to support credit analysts, relationship managers, and committees with consistent and transparent recommendations.</p><p>This is a simple starting point, but it shows how AI can make credit decisions better, faster, and more reliable.</p><h3>What is Deep Learning?</h3><p>Before we dive into the business use cases and technical aspects of this project. Let me share a brief introduction to deep learning.</p><p>D<strong>eep learning</strong> is a subfield of AI that mimics how the human brain works.<br> Just like the brain has billions of neurons passing signals, deep learning models have <strong>layers of “neurons”</strong> — called nodes — that pass <strong>numerical signals</strong> (called activations) between them.</p><p>These models <strong>learn</strong> by adjusting the <strong>strength of connections</strong> (called weights) between nodes, similar to how our brain reinforces certain connections as we learn new skills.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gRFcLCEOEKk4SRt1eVHKuQ.png" /><figcaption>Sample neural network diagram from IBM deep learning course</figcaption></figure><p><strong>Image of AI neurons vs biological neurons — </strong>showcasing how deep learning tries to mimic the human brain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/610/0*fQjqBuzWYMyGHM7l.png" /><figcaption>Image Credit: orignal creator</figcaption></figure><h3>What’s the math behind it?</h3><p>Under the hood, it’s a combination of:</p><ul><li><strong>Linear algebra</strong>: multiplying inputs and weights</li><li><strong>Non-linear functions</strong>, like ReLU or sigmoid to simulate brain-like firing</li><li><strong>Loss functions</strong>: which measure how wrong the prediction is</li><li><strong>Gradient descent</strong>: a way to minimize error step-by-step, like slowly improving your guesses</li></ul><h4>Important Terms</h4><p>This project uses Keras to build simple deep learning models using neural network using <strong>ReLU</strong> activation and <strong>adam</strong> optimizer.</p><p><strong>ReLU (Rectified Linear Unit)</strong> activation function outputs the input if it’s positive, and outputs zero if the input is negative.</p><p><strong>adam</strong> optimizer is used for updating network weights during training. It’s a gradient-based optimization algorithm that adapts the learning rate for each parameter individually, making it efficient and robust.</p><p><strong>Epochs</strong> — It’s a fundamental concept used to measure the progress of learning. It represents one cycle of training where the model processes all the data in the dataset.</p><h3>Business Objective</h3><p>Corporate banks often provide credit lines to businesses to help them manage liquidity and working capital. However, assigning an optimal credit line involves risk assessment, cash flow forecasting, and client profiling.</p><p>This project aims to use deep learning to <strong>predict the optimal credit line</strong> for a corporate client. The goal is to help banks:</p><ul><li><strong>Improve capital efficiency</strong></li><li><strong>Reduce credit risk</strong></li><li><strong>Enhance client satisfaction with tailored credit offers</strong></li></ul><h4>Key Business Questions</h4><ol><li><strong>How can we predict the right credit line for a business given its financial and operational data?</strong></li><li><strong>Can AI models outperform heuristic-based or rule-based credit allocation?</strong></li><li><strong>How can predictive models reduce NPLs (Non-Performing Loans) while maximizing revenue?</strong></li></ol><h4>Target Users</h4><ul><li><strong>Credit Risk Analysts</strong>: Use predictions as part of credit evaluation workflows.</li><li><strong>Relationship Managers</strong>: Provide data-backed advice on credit lines to corporate clients.</li><li><strong>Credit Committee Members</strong>: Use model output to support or challenge human credit assessments.</li></ul><h4>Strategic Business Benefits</h4><ul><li><strong>Risk Management </strong>— Models can reduce overexposure by suggesting more conservative limits for risky clients. However, this can impact the revenue. We will need to take a balanced approach as per the organisation&#39;s risk appetite.</li><li><strong>Client Retention </strong>— Underserved clients with solid fundamentals can be offered better credit, improving satisfaction.</li><li><strong>Operational Efficiency</strong> — Speeds up credit decision-making with consistent, data-backed recommendations.</li><li><strong>Regulatory Compliance:</strong> We must choose a model that provides transparency and explainability rather than pure accuracy performance metrics. Tools like SHAP can explain and audit models, aiding compliance with Basel III/IV.</li></ul><h4>Business Constraints &amp; Considerations</h4><ul><li><strong>Explainability</strong>: Corporate lending decisions must be explainable for compliance and internal governance.</li><li><strong>Bias and Fairness</strong>: Must ensure the model does not systematically disadvantage certain industries or regions.</li><li><strong>Data Quality</strong>: Corporate financials can vary in structure and reliability; preprocessing is key.</li><li><strong>Model Governance</strong>: Model decisions should not override human judgment but support it.</li></ul><h4>Success Metrics</h4><ul><li><strong>MAE or RMSE</strong>: For model performance.</li><li><strong>Reduction in credit loss (%)</strong>: Compared to traditional underwriting.</li><li><strong>Approval turnaround time (TAT)</strong>: Time saved in credit line decision-making.</li><li><strong>Client NPS</strong>: Improvement in Net Promoter Score due to fairer, faster credit offers.</li><li><strong>Model Explainability</strong> using SHAP</li></ul><h3>Tech Stack</h3><p>This section provides a technical stack overview used in this project.</p><ul><li><strong>Language</strong>: Python</li><li><strong>ML Framework</strong>: Deep Learning using Keras</li><li><strong>Model Type</strong>: Regression (e.g., Dense Neural Network)</li><li><strong>Tools</strong>: pandas, numpy, scikit-learn (for preprocessing), SHAP (for explainability)</li><li><strong>Data source</strong>: <a href="https://www.kaggle.com/datasets/agewerc/corporate-credit-rating/data">Kaggle</a> data with rich financial indicators for corporate clients</li><li>Github: <a href="https://github.com/dhanmoti/ml-article-sample-code/blob/main/EDA.ipynb">https://github.com/dhanmoti/ml-article-sample-code/blob/main/EDA.ipynb</a></li><li><strong>IDE</strong>: Visual Studio</li></ul><h4><strong>Model Implementation</strong></h4><p><strong>The project structure of this project</strong></p><pre>credit_line_ai/<br>├── data/<br>│   ├── raw/                     # Original CSV/Excel data<br>│   └── processed/               # Cleaned datasets<br>├── notebooks/<br>│   └── EDA.ipynb                # Exploratory Data Analysis<br>├── models/<br>│   └── credit_model.h5          # Saved Keras model<br>├── src/<br>│   ├── data_preprocessing.py<br>│   ├── train_model.py<br>│   ├── predict.py<br>│   └── explain_model.py<br>└── requirements.txt</pre><p>The code below creates a very simple regression model using Keras Sequential model. It has one input layer, 5 hidden layers, and one output layers. In real world projects, we will have many layers. Each hidden layer has 50 neurons. Fit function here, train and test the model at the same time using 100 epochs.</p><pre>def regression_model():<br>    input_colm = predictors_norm.shape[1] # Number of input features<br>    # create model<br>    model = Sequential()<br>    model.add(Input(shape=(input_colm,)))  # Set the number of input features <br>    model.add(Dense(50, activation=&#39;relu&#39;))  <br>    model.add(Dense(50, activation=&#39;relu&#39;))<br>    model.add(Dense(50, activation=&#39;relu&#39;)) <br>    model.add(Dense(50, activation=&#39;relu&#39;))<br>    model.add(Dense(50, activation=&#39;relu&#39;))  <br>    model.add(Dense(1))  # Output layer<br>    <br>    # compile model<br>    model.compile(optimizer=&#39;adam&#39;, loss=&#39;mean_squared_error&#39;)<br>    return model</pre><pre># fit the model<br>model.fit(predictors_norm, target, validation_split=0.3, epochs=100, verbose=2)</pre><p><strong>Training Process</strong></p><p>Depending on the amount of data, this process can be very slow. As you can see in the screenshot below, train-loss and validation-loss Improves after each epoch cycle. We can further improve this model by adding more hidden layers, more neurons to improve the model&#39;s performance. Please note that we will need to keep an eye on model such that it is neither underfit nor overfit. But model evaluation will not be covered in this article.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Y1mnn2Ix1tC3tTksNiZ4OA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/868/1*TEp2RJAe7zcne6E4LZ1Q6w.png" /><figcaption>SHAP Analysis of trained model</figcaption></figure><p><strong>Model Deployment</strong></p><p>Save the model using the following code so that we can develop an API using the Flask library.</p><pre># Save model <br>model.save(MODEL_PATH)<br>joblib.dump(scaler, os.path.join(&quot;models&quot;, &quot;scaler.pkl&quot;))<br>print(&quot;Model and scaler saved.&quot;)</pre><p><a href="https://www.linkedin.com/pulse/creating-api-simple-linear-regression-model-using-python--k5khc">Creating an API for a Simple Linear Regression Model Using Python</a></p><h3>Addressing Key business questions</h3><h4>1. How can we predict the right credit line for a business given its financial and operational data?</h4><p>Model estimates <strong>credit risk</strong> from a rich set of financial ratios and SHAP values explain which factors most influence that risk. From here:</p><ul><li>Businesses with <strong>lower predicted credit risk</strong> can be granted <strong>higher credit lines</strong>.</li><li>A sample formula that converts the <strong>predicted risk score</strong> to a <strong>recommended credit line</strong></li></ul><pre>Credit Line = f(1 / Predicted Risk, Cash Flow Health, Asset Turnover)</pre><p>SHAP analysis helps <strong>justify</strong> the credit decision: e.g., “This business has strong operating margins and low debt, so we recommend a higher line.”</p><p>2. <strong>Can AI models outperform heuristic-based or rule-based credit allocation?</strong></p><p>Yes in long run — here’s how an AI model adds value:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fc4TXjbNBKxuy06ZKqCD2A.png" /></figure><p>3. How can predictive models reduce NPLs (Non-Performing Loans) while maximizing revenue?</p><p>Model helps optimize the <strong>risk–return trade-off</strong>:</p><p><strong>Lower NPLs</strong>:</p><ul><li>By flagging businesses at higher risk of default</li><li>Allowing early intervention, restructured terms, or outright rejection</li></ul><p><strong>Maximized revenue</strong>:</p><ul><li>Approving low-risk clients who may have been missed by rigid rules</li><li>Right-sizing the credit line: offering more to good borrowers</li></ul><p><em>In essence, the model enables </em><strong><em>risk-based pricing and limit-setting</em></strong><em> — charge higher rates to riskier clients, and offer better terms to reliable ones.</em></p><h3>Implementation Strategy</h3><h4>Transition Vision</h4><p><strong>From: </strong>Manual, rule-based, or spreadsheet-driven credit line allocation based on human judgment and basic financial ratios.</p><p><strong>To: </strong>AI-assisted decision support system that uses historical data and financial indicators to recommend optimized credit limits, integrated into the corporate credit workflow.</p><h4>Roadmap</h4><p>If the organisation is already practicing Agile, adopting rapid iteration methods is easier. Otherwise, I recommend a mix of parallel and phased implementation approaches. Although parallel implementation is one of the most expensive approaches, considering the risk involved, it is best suited for our early-stage model in the lending business.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QOLCjAwtwGnf0Lo00o-50A.png" /><figcaption>Sample roadmap for organisation to adopt to new AI powered Credit Line Optimization system</figcaption></figure><p><strong>Potential Stakeholders</strong></p><p>List of potential stakeholders and their roles:</p><ul><li>Credit Team — Co-design decision rules and overrides</li><li>IT/Data Engineering — Build pipelines and deploy models to existing systems</li><li>Risk &amp; Compliance — Validate model behavior and fairness</li><li>Senior Management — Endorse and drive change adoption</li><li>End Users (RMs, Analysts) — Receive training and provide feedback</li></ul><p><strong>Tools &amp; Infrastructure Needed</strong></p><p>This section needs to be monitored and revised as the product evolves.</p><ul><li><strong>ML Platform</strong>: Python/Keras for prototype</li><li><strong>Database Access</strong>: To ingest financials and outcomes.</li><li><strong>Explainability Libraries</strong>: SHAP or LIME for risk oversight.</li><li><strong>Monitoring Tools</strong>: Model performance, data drift, usage logs.</li></ul><h3>Conclusion</h3><p>This project is a small but meaningful step toward using deep learning to improve decision-making in business. By building a basic neural network with Keras, we can predict credit lines.</p><p>While the model is simple, it highlights the value of using data and machine learning to support better, faster, and fairer lending. With further development, this approach can become a powerful tool for credit teams — boosting decision quality, reducing risk, and improving the overall client experience.</p><p>Thank you for reading till here! 🙏✨ Keep innovating!</p><p>Ideated and edited by ChatGPT.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=895ae365bf0d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[YAGNI: You Aren’t Gonna Need It — A Developer’s Guide]]></title>
            <link>https://medium.com/@dhanmoti/yagni-you-arent-gonna-need-it-a-developer-s-guide-bb05c4a3aa89?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/bb05c4a3aa89</guid>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Fri, 14 Feb 2025 14:55:12 GMT</pubDate>
            <atom:updated>2025-02-14T14:55:12.138Z</atom:updated>
            <content:encoded><![CDATA[<h3>YAGNI: You Aren’t Gonna Need It — A Developer’s Guide</h3><p>YAGNI is a core tenet of agile and extreme programming that advises against writing code or building features until they’re absolutely required.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*02tMZdkFPHT6p1pwlH_TuQ.jpeg" /><figcaption>Image generate by ChatGPT</figcaption></figure><h3>Impact on business when YAGNI principle is violated</h3><p>Last week’s dispute role-play exercise at my MBA studies reminded me of this principle.</p><p>While YAGNI is a principle rooted in software development, its impact reaches far beyond just code and architecture. For those in business management, understanding and applying YAGNI can lead to significant improvements in project management, cost efficiency, and time-to-market. Let’s take a hypothetical example of a project where conflicting priorities among stakeholders caused delays, resulting in wasted resources and tension within the team.</p><p>In a typical project, a lead developer, perhaps confident in their skills and vision, might decide to over-engineer a system by adding features or building for potential future needs — features that aren’t required by the client or the project in its current scope. This situation can create a ripple effect: deadlines slip, resources are diverted, and costs escalate.</p><p>This situation mirrors what happens in businesses when excessive planning for the future leads to the accumulation of “unnecessary” features or decisions. It detracts from delivering what the customer actually wants, creating more problems than it solves.</p><p><strong>Why YAGNI Helps Avoid These Pitfalls:</strong> By adopting YAGNI, businesses can:</p><ul><li><strong>Reduce Wasted Resources</strong>: Resources (time, effort, money) aren’t spent on features that may never be used. Instead, companies can focus on delivering immediate value to clients.</li><li><strong>Avoid Scope Creep</strong>: YAGNI keeps the focus on what is truly necessary right now, preventing projects from expanding into areas that dilute focus and increase complexity.</li><li><strong>Align Team Priorities</strong>: When teams are aligned around delivering only what’s required for the immediate task, it prevents miscommunications and misunderstandings between different stakeholders, leading to more harmonious team dynamics.</li></ul><h3>Intro</h3><p>YAGNI is a core tenet of agile and extreme programming(XP) that advises against writing code or building features until they’re absolutely required. <strong><em>In short, don’t implement functionality until you’re certain it’s needed.</em></strong> By doing so, you save time, reduce complexity, and let your project evolve naturally through real user feedback instead of assumptions.</p><h4>Metaphor</h4><p>Imagine you’re opening a restaurant. When planning your kitchen, you might be tempted to purchase every fancy appliance — state-of-the-art ovens, sous-vide machines, gourmet espresso makers — anticipating future expansion. However, if your customers are initially happy with a simple, well-prepared menu, investing in high-end equipment becomes an unnecessary expense.</p><p>In programming, building only the essential features is like equipping your kitchen with just the basics. You deliver what’s needed now, and as customer demands evolve, you can add extra features gradually. This avoids the clutter and maintenance overhead associated with speculative, over-engineered code.</p><h4>What is YAGNI?</h4><blockquote>“You Aren’t Going to Need It: Now or Later?<br>We are traditionally told to plan for the future … Instead, XP says to do a good job … of solving today’s job today … and add complexity in the future where you need it. The economics of software as options favor this approach. [Beck1999]</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t17iP1yxM0zPg4ixRM6ijQ.png" /><figcaption>Source: <a href="http://www.geeksforgeeks.org/">www.geeksforgeeks.org</a></figcaption></figure><blockquote>One of the most widely publicized principles of XP is the You Aren’t Going to Need It (YAGNI) principle. The YAGNI principle highlights the value of delaying an investment decision in the face of uncertainty about the return on the investment.” — Beck, K. (1999)</blockquote><p>This quote underscores that focusing on today’s needs can offer both economic and practical benefits.</p><p>Robert C. Martin, in <em>Clean Code</em>, reinforces this principle, emphasizing that <strong>the best code is simple and minimal</strong>. He argues that writing unnecessary code increases <strong>complexity and maintenance costs</strong>, ultimately violating one of the fundamental principles of clean software craftsmanship (Martin, 2008).</p><p>Martin Fowler expands on this by emphasizing that <em>YAGNI is about avoiding unnecessary investments</em>. He argues that developers often believe they are “saving time in the future” by preemptively adding extra features, but in reality, <strong>this leads to increased costs and maintenance burdens because requirements frequently change</strong> (Fowler, 2004).</p><h3>Navigating a Challenging Leadership Environment</h3><p>If you find yourself as a junior programmer in a situation where you are tasked to implement future requirements, consider the following steps:</p><ol><li><strong>Implement and Document Thoroughly:</strong><br>Follow your lead’s instructions and implement the extra features, but document every detail. Keep clear records of emails, meeting notes, and commit messages that capture the directions and your interpretation of the task. This documentation provides evidence that you followed orders.</li><li><strong>Clarify in Writing:</strong><br>Politely request written clarifications. For example, ask, “To confirm, you’d like the extra functionality added now even though it extends beyond our core requirements?” This creates a paper trail that can help if issues arise later.</li><li><strong>Use Feature Flags and Modular Design:</strong><br>Where feasible, implement extra features behind feature flags or as modular components. This design approach allows for easy disabling or removal later if the extra functionality proves unnecessary. Assuming you have some control over software architecture and code design patterns.</li><li><strong>Maintain a Minimal Branch:</strong><br>Consider keeping a separate branch or commit that contains only the core, minimal implementation. This backup is invaluable if the project’s direction changes or if your additional work becomes a liability.</li><li><strong>Back-Up Your Repository:</strong><br>Take screenshots or maintain a backup of the remote repository. Given that version control operations like git-rebase can rewrite commit history, having a saved snapshot helps document what was originally implemented.</li><li><strong>Seek Mentorship and Peer Advice:</strong><br>Discuss your concerns with a mentor or a trusted colleague. Their insights can help you navigate the situation tactfully while meeting the lead’s demands and adhering to best practices.</li></ol><p>These measures help you protect your professional integrity while fulfilling your assigned tasks, even under challenging leadership dynamics.</p><h3>When to Use YAGNI</h3><ul><li><strong>Agile Environments:</strong> When requirements are rapidly evolving, YAGNI keeps you focused on delivering working software quickly.</li><li><strong>Prototyping and MVPs:</strong> During early project stages, implement only what’s essential to validate your ideas, minimizing wasted effort.</li><li><strong>Reducing Technical Debt:</strong> Avoiding speculative features helps maintain a cleaner codebase, reducing maintenance challenges over time.</li></ul><h3>When Not to Use YAGNI</h3><ul><li><strong>Stable, Well-Defined Requirements:</strong> If the long-term roadmap is clear and future needs are predictable, planning ahead can be justified.</li><li><strong>Core Infrastructure Components:</strong> For foundational frameworks or libraries that support multiple projects, a degree of foresight is beneficial.</li><li><strong>High Cost of Deferred Work:</strong> In cases where postponing a feature might necessitate significant refactoring later, some level of planning is warranted.</li></ul><h3>Conclusion</h3><p>The YAGNI principle is a powerful reminder that over-engineering can lead to wasted resources, delayed delivery, and quality issues. Much like a restaurant that starts with only essential kitchen equipment and scales up based on demand, a codebase built on YAGNI remains lean and adaptable. While challenging leadership dynamics — such as those observed in our MBA dispute mediation role play — can complicate its application, following best practices of software programming like thorough documentation, clear written clarifications, modular design, and repository backups can help protect your work.</p><blockquote>Happy coding !!!</blockquote><h3>References</h3><ul><li>Beck, K. (1999). <em>Extreme Programming Explained: Embrace Change</em>. Addison-Wesley.</li><li>Martin, R. C. (2008). <em>Clean Code: A Handbook of Agile Software Craftsmanship</em>. Prentice Hall.</li><li>GeeksforGeeks. (n.d.). What is YAGNI Principle? You Aren’t Gonna Need It. Retrieved from<a href="https://www.geeksforgeeks.org/what-is-yagni-principle-you-arent-gonna-need-it/"> https://www.geeksforgeeks.org/what-is-yagni-principle-you-arent-gonna-need-it/</a></li><li>Beck, K. (1999). <em>Extreme Programming Explained: Embrace Change</em>. Addison-Wesley.</li><li>VersionOne. (n.d.). 15th Annual State of Agile Report. Retrieved from <a href="https://stateofagile.com/">https://stateofagile.com/</a></li><li>Forsgren, N., Humble, J., &amp; Kim, G. (2018). <em>Accelerate: The Science of Lean Software and DevOps</em>. IT Revolution Press.</li></ul><p>Edited by ChatGPT</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bb05c4a3aa89" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Dark Side of Git Rebase: When It Becomes a Tool for Disruption]]></title>
            <link>https://medium.com/@dhanmoti/the-dark-side-of-git-rebase-when-it-becomes-a-tool-for-disruption-00be2846135a?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/00be2846135a</guid>
            <category><![CDATA[teamwork]]></category>
            <category><![CDATA[psychological-safety]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[trust]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Tue, 11 Feb 2025 13:48:33 GMT</pubDate>
            <atom:updated>2025-02-11T13:51:18.413Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>The Hidden Risks of Git Rebase in a Collaborative Team</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tiFIvZnjzArj7wS1cGippw.jpeg" /><figcaption>Photo credit to original owner (<a href="https://unsplash.com/@yancymin">Yancy Min</a>)</figcaption></figure><p>In software development, Git is an essential tool for version control, helping teams collaborate efficiently. One of its most powerful features — <strong>Git rebase</strong> — is often misunderstood and, in some cases, misused. While rebase can streamline history and keep commit logs clean, it also has the potential to disrupt teamwork when applied carelessly. Worse, when used maliciously, it can erode trust within a team.</p><h3>What is Git Rebase and When to Use It?</h3><p>Git rebase is a way to integrate changes from one branch onto another by rewriting commit history. Unlike merging, which creates a new commit to reflect integration, rebasing applies changes sequentially on top of a target branch, maintaining a linear history.</p><h3>When Rebase is Beneficial:</h3><ul><li>Keeping feature branch history clean before merging.</li><li>Avoiding unnecessary merge commits in personal development branches.</li><li>Ensuring a smooth integration when collaborating with upstream repositories.</li></ul><p>However, <strong>rebasing a shared branch, especially with a force push, can be risky</strong> because it rewrites commit history for everyone working on the branch.</p><h3>The Dark Side of Git Rebase: When It Becomes a Tool for Disruption</h3><p>Imagine working on a shared development branch, collaborating with your team. Suddenly, without warning, the branch is rebased and force-pushed. Your local history no longer aligns with remote, causing conflicts, lost commits, and wasted hours of debugging.</p><p>Now, consider a scenario where this isn’t just a one-time mistake but a recurring issue. A team member rebases and force-pushes changes without prior communication, making collaboration difficult. Worse, if someone deliberately modifies commit history, it can be used to obscure changes, shift blame, or manipulate records — damaging team trust.</p><p>This highlights an <strong>often-overlooked risk in version control: the potential for rebase to be weaponized in a toxic work environment.</strong> Without safeguards, bad actors can exploit Git’s flexibility to fabricate commit history, making accountability difficult.</p><h3>Best Practices to Avoid Rebase-Related Conflicts</h3><p>To ensure rebase is used productively and not destructively, teams should establish clear guidelines:</p><p><strong>Communicate Before Rebasing Shared Branches</strong></p><ul><li>Notify team members before performing a rebase on a collaborative branch.</li><li>Use team standups, Slack messages, or a quick heads-up before rewriting history.</li></ul><p><strong>Protect Critical Branches from Force Pushes</strong></p><ul><li>Enable branch protection rules in GitHub, GitLab, or Bitbucket.</li><li>Require pull requests and approvals before merging significant changes.</li></ul><p><strong>Use Signed Commits for Accountability</strong></p><ul><li>Encourage cryptographic signing of commits to verify authorship.</li><li>This prevents unauthorized tampering and adds an extra layer of integrity.</li></ul><p><strong>Document Changes Transparently</strong></p><ul><li>When rebasing is necessary, maintain clear documentation.</li><li>Use git reflog to track history in case of accidental rewrites.</li></ul><p><strong>Foster a Culture of Trust and Psychological Safety</strong></p><ul><li>Engineering teams should create an environment where collaboration and respect come first.</li><li>If someone repeatedly disrupts the workflow, leadership should address the issue rather than dismiss concerns.</li></ul><h3>Final Thoughts: Rebase with Responsibility</h3><p>Git rebase is a powerful tool, but with great power comes great responsibility. While it can enhance efficiency, reckless or malicious rebasing can harm both productivity and team morale. Organizations must recognize the importance of psychological safety in engineering teams and ensure that technical tools are used to build trust, not break it.</p><p>Have you ever faced challenges with Git rebase in a team setting? Share your thoughts and experiences in the comments!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S_r_lAbRdqZgY_GgnnSNPQ.jpeg" /></figure><p>👩‍💻 🧑‍💻 Happy coding !!!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=00be2846135a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Modularisation in Software Engineering]]></title>
            <link>https://medium.com/@dhanmoti/modularisation-in-software-engineering-4435500437dd?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/4435500437dd</guid>
            <category><![CDATA[software-modularity]]></category>
            <category><![CDATA[modularization]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[software-engineering]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Tue, 17 Oct 2023 14:55:29 GMT</pubDate>
            <atom:updated>2023-10-17T14:57:22.516Z</atom:updated>
            <content:encoded><![CDATA[<p>Modularisation is popular among programmers these days. What is modularisation?</p><h3>What is modularisation?</h3><p>First let’s answer <strong><em>What is Software Architecture?</em></strong></p><blockquote>“is to facilitate the <strong>development, deployment, operation and maintenance</strong> of the software system contained within it” — Clean Architecture by Robert C. Martin</blockquote><p>With this definition of Architecture in mind, lets talk about modularisation.</p><p>Modularisation means separating or grouping codes into smaller entities/components so that source code is reusable, testable, and scalable.</p><h3>Tradeoff between number of modules and maintenance complexity</h3><p>Everybody is recommending…Modularize, Modularize, Modularize… but how many modules is best for your project and team?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zxkdEgEP3RvHHBpJTk5Rng.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8Ks1uqGZ-bxxJqvYUHkD1w.png" /><figcaption>Research Article: Towards an Analytical Approach to Measure Modularity in Software Architecture Design</figcaption></figure><h4>How to find the right number of modules for a particular project?</h4><p>I do not know of any formula. But one way is to focus on problem at hand and number of developers working on the same codebase.</p><h4>Feature related issue</h4><p>Let say we have two similar touring applications one is for tourist visiting new places and another one is for school children learning inside a botanical garden. Both applications use AR and geofencing technology to prompt related information regarding their surroundings. Problem is <em>developers need to write AR and geofencing related codes in both applications</em>. Extracting out these features reduce development effort for a team of 1–2 developers.</p><h4>Development team size related issue</h4><p>Image you have 2 developers. Number of components/modules is 10–20. To finish on feature one developer have to write code in ten different modules. Make sure these modules work as expected when combine together. Most of the time, developers will be busy writing codes that link these modules. But, when development team is mid-large(50–100), <em>it is important to have clear boundary and modules so that large team can work together.</em> It is even more helpful when these software have more than 5-10 main features. Most of these features are shared in multiple different applications.</p><h3>How NOT to modularise</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_2lSNwYHmukF3pFz05Nl-A.png" /><figcaption>Sample High Level Architecture (How NOT to modularise)</figcaption></figure><h4>How this fails to meet the original goal of modularisation?</h4><p>Goal of modularisation for this project is to draw clear boundary between modules. For example, when Module Bs import Module A, it is should NOT see other modules related class/interface/protocol. It should hide its details.</p><ol><li>As every modules puts their interface/protocol inside Module D and almost everyone import Module D to implement their related interface/protocols, every module can see all the interface/protocols used in the entire project.</li><li>All module make their concrete classes public so that those who need to create these classes can access. It does not use any creation design patterns inside its modules. It is does not hide any detail implementations.</li></ol><p>One easy way to fix is have a dedicated abstract module instead of putting all abstract classes in one module. To hide its detail, one could use <a href="https://refactoring.guru/design-patterns/factory-method">Factory Design patterns</a>. Make Factory layer public, not concrete classes.</p><p>We should also ask - Do we really need abstract module for every module? What is the benefit? How many different implementation are already there in our project? what is the likelihood of implementing additional classes for same abstract classes?</p><h3>How to modularise?</h3><blockquote>Having swift package manager is NOT (NOT) modularisation!</blockquote><p>It makes the rules part of IDE. It makes it easier to share across applications. <strong><em>If the architecture is wrong in conceptual level, SPM cannot save the project.</em></strong></p><h4>Ways to Modularise</h4><p>There are multiple ways to achieve this goal. I find these two useful for small applications with 2–3 developers .</p><ul><li>Layer approach: Separating based on dependencies such as networking, repository, user interface, business logic. Grouping by similarity such as shared user interfaces, utilities. Easy to implement. This approach cannot share feature across multiple different projects.</li><li>Feature approach: Separating based on releasable components such as feature A, feature B etc. Easy to over modularised. If not done well, project might reach its peak complexity in much shorter time. This approach allows developer to build multiple applications that uses similar features.</li></ul><p>Both approach come with its benefits and limitations. <strong><em>We need to tradeoff between complexity and reusability.</em></strong> Do you really need to separate this layer right now or partial border is good enough for now?</p><p>It is good to get the conceptual level right first. Just a simple diagram with few boxes is good enough. Don’t jump to your keyboard and starting writing code. Later it is harder to modify. Get feedback from team or other developers as early as possible. Document concept and rules set for modularisation. Share it with team so that your team is on the same page.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lfy7HLSUCjvmhilOZ3ysUQ.png" /><figcaption>Layer Approach (We can also extract out UI Navigation as one more layer but again do we really need it now?)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mze2R2bquJVfDlZB7z7_Qw.png" /><figcaption>Layer + Feature Approach (Here extracting out UI Navigation can increase reusability of Feature but will add extra class to implement and test)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e0uLI-_8cgQWNVdHdu0YHQ.png" /><figcaption>Feature Approach</figcaption></figure><h4>References</h4><p><a href="https://www.researchgate.net/publication/283201861_Towards_an_Analytical_Approach_to_Measure_Modularity_in_Software_Architecture_Design">https://www.researchgate.net/publication/283201861_Towards_an_Analytical_Approach_to_Measure_Modularity_in_Software_Architecture_Design</a></p><p><a href="https://dev.to/prxtikk/how-to-write-clean-and-modular-code-1d87">How to write Clean and Modular Code 🤩👨‍💻</a></p><p><a href="https://developer.android.com/topic/modularization">https://developer.android.com/topic/modularization</a></p><p><a href="https://developer.android.com/topic/modularization/patterns">https://developer.android.com/topic/modularization/patterns</a></p><blockquote>“Any fool can write code that a computer can understand. Good programmers write code that humans can understand.” <em>— </em><strong><em>Martin Fowler</em></strong></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4435500437dd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Coding | Meaningful Names]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@dhanmoti/intention-revealing-names-765ac3427cf8?source=rss-34d60b1d76d1------2"><img src="https://cdn-images-1.medium.com/max/600/1*Vfcy_SZyerrqPIIdcpbzVQ.png" width="600"></a></p><p class="medium-feed-snippet">Notes from &#x201C;Clean Code&#x201D; by Robert C. Martine. | Chapter 2: Meaningful Names</p><p class="medium-feed-link"><a href="https://medium.com/@dhanmoti/intention-revealing-names-765ac3427cf8?source=rss-34d60b1d76d1------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@dhanmoti/intention-revealing-names-765ac3427cf8?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/765ac3427cf8</guid>
            <category><![CDATA[clean-code]]></category>
            <category><![CDATA[programmer-life]]></category>
            <category><![CDATA[coding-best-practices]]></category>
            <category><![CDATA[coding-notes]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Wed, 22 Dec 2021 03:38:46 GMT</pubDate>
            <atom:updated>2021-12-23T08:46:38.835Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Clean Code — Why care?]]></title>
            <link>https://medium.com/@dhanmoti/clean-code-why-care-e22bf6bf1388?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/e22bf6bf1388</guid>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[clean-code]]></category>
            <category><![CDATA[developer]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Sun, 06 Oct 2019 14:43:23 GMT</pubDate>
            <atom:updated>2019-10-06T14:43:23.807Z</atom:updated>
            <content:encoded><![CDATA[<h3>Clean Code — Why care?</h3><p>Chapter 1: Clean Code</p><p>Recently, I bought this book called “Clean Code” by Robert C. Martine. For the first time, I see programming under a different light. Programmers are authors here who write in such a way that even dummy can understand.</p><p>This is the summary of Chapter 1 which explains <strong>why one should care about clean code</strong>.</p><p>Good news for programmers…</p><blockquote>“We will never be rid of code. Code is specification of requirements in such detail that a machine can execute them.” — Robert C. Martine</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iLAXHkr0mH8uiJFszL4dfA.jpeg" /></figure><h3>Bad Code: A sad story of a killer app</h3><p>In the late 80s, one company wrote this killer app which was very popular, and lost of professionals bought and used it. But then…</p><p>The release cycles began to stretch.</p><p>Bugs were not repaired from one release to the next.</p><p>Load times grew and crashes increased.</p><p>And…</p><p>The product was shut down and the company went out of business a short time after that.</p><p>How did this happen? — they had rushed the product and made a huge mess in the code. As more and more features were added, the code got worse and worse until it simply became unmanageable.</p><blockquote>“It was the bad code that brought the company down.”</blockquote><h3>Cost of Owning a Mess</h3><p>As the mess builds, the productivity of the team continues to decrease and the management add more staff in hopes of increasing productivity. With the <strong>horrific pressure to increase productivity</strong>, they all make more and more messes, <strong>driving the productivity ever further toward zero</strong>.</p><h3>Attitude</h3><p>I learnt that we could easily blame management who try to push deadlines or seniors who created the mess. But if we continue keeping this mess, it is so unprofessional of us.</p><p>I myself once complained about a code, which should be one-line change made instead in hundreds of different places, to my seniors and even management. When they denied to improve I continued creating the mess like others. As a result, I also needed to do overtime like others to meet the deadlines. This would had further decreased overall productivity of entire team.</p><h3>Conundrum</h3><blockquote>“Previous messes slow programmers down. Yet all the pressure to make messes in order to meet deadlines.” In short, we never take time to go fast!</blockquote><p>As true professionals, we should avoid the second part of the conundrum at all cost.</p><h3>So…What is clean code?</h3><p>“The logic should be straightforward to make it hard for bugs to hide. Clean code does one thing well.” — Bjarne Stroustrup, inventor of C++</p><p>“Clean code reads like well-written prose.” — Grady Booch, author of OOAD with Applications</p><p>“Clean code has meaningful names. It provides on way rather than many ways for doing one thing.” — [Big] Dave Thomas, founder of OTI, godfather of the Eclipse strategy</p><p>“…when each routine you read turns out to be pretty much what you expected.” — Ward Cunningham, inventor of Wiki.</p><h3>We Are Authors</h3><blockquote>“Responsible for communicating well with our readers.”</blockquote><p>This is the end of Chapter 1. Next is about Names. I will update the link here once summary of Chapter 2 is available. Thanks for ready till here.</p><p>Happy coding!</p><p>Do remember not to create mess in order to meet deadlines!!! 😇</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e22bf6bf1388" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Web App Idea: Exchange Rates around Yangon]]></title>
            <link>https://medium.com/@dhanmoti/web-app-idea-exchange-rates-around-yangon-697e0a157a07?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/697e0a157a07</guid>
            <category><![CDATA[myanmar-kyat]]></category>
            <category><![CDATA[finance]]></category>
            <category><![CDATA[web-app-development]]></category>
            <category><![CDATA[exchange-rate]]></category>
            <category><![CDATA[mmk]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Mon, 04 Feb 2019 08:14:47 GMT</pubDate>
            <atom:updated>2019-02-04T12:50:43.065Z</atom:updated>
            <content:encoded><![CDATA[<h3>Users</h3><h4>Browse Exchange Rates</h4><ul><li>Central Bank</li><li>Banks ( KBZ, CB, Aya Bank, Myanma Foreign Trades, United Amara Bank)</li><li>Top Picks by Developer (9 money changers)</li><li>Show buy/sell rates with specific notes if available</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/903/1*wslFkG3IBMJ5rL1fjNnVcw.png" /><figcaption>MMK Exchange Rate tracker logo</figcaption></figure><h4>Find money changers</h4><ul><li>by currency (not all hv all the currencies)</li><li>Name</li><li>Location</li><li>Type: Bank, shop, finance company</li></ul><h4>Track Currencies</h4><ul><li>Central bank only for now</li></ul><h4>Suggest Money Changers</h4><ul><li>new money changers</li><li>new bank with has currency exchange service</li></ul><h3>Money Changers</h3><ul><li>Sign Up (with logo, brand message)</li><li>Post daily Exchange Rates of FB page</li><li>Widget for websites</li><li>Console/Dashboard to update latest rates</li></ul><h3>Version 1 — Daily Rates Updating Process</h3><h4>Daily Updates for Top Picks Money Changers and Top 3 Banks</h4><p>- Manually update rates for now</p><p>- Allow them to add new money changers (only Yangon Region) up to 30 money changers.</p><p>- Review new money changes and active their pages on mmk tracker (by Developer)</p><p>- Allow interested money changers to signup (upon sign ups, open up rates updaters to respective money changers)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=697e0a157a07" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Transferring Money using CB Mobile app in Myanmar(Burma)]]></title>
            <link>https://medium.com/@dhanmoti/transferring-money-using-cb-mobile-app-in-myanmar-burma-fe5e2e4418c0?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/fe5e2e4418c0</guid>
            <category><![CDATA[myanmar]]></category>
            <category><![CDATA[ui]]></category>
            <category><![CDATA[mobile-app-development]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Sat, 24 Nov 2018 16:02:40 GMT</pubDate>
            <atom:updated>2019-02-18T12:37:16.127Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Background</strong></h3><p>It all started at the Changi Airport. I was going back to Yangon for a family trip and I had promised that I will take care of the budget. Guess what happened? I could not withdraw cash at airport due to my withdrawal limits. I was very disappointed and landed in Yangon without enough cash for the trip. My parents picked me up from the airport. I had to wait for my sister to come home. It was weekday, she was working.</p><p>Meanwhile, I remembered there was a feature which allows us to <strong>transfer money from DBS to CB Bank</strong>. So, I give it a try. Woowww!!!!, just before the trip began, I got the money into my Myanmar bank account. But my sister said she will lead the trip and we settle the amount later. It was like she pay the 50% and I pay 50%. So I agreed.</p><p>At the end of the trip, when my sister was ready with the expense details and how much i needed to pay her, I took out my phone and logged into CB Bank Mobile App. It went something like this.</p><h3>Problem</h3><ol><li>I asked her account number which she sent me a screenshot of that.</li></ol><blockquote>I tried to switch between photos app and CB Mobile App, and I was logged out when I went back to CB Mobile App.</blockquote><blockquote>— Isn’t it too much? If security is the concerns, it can be done in many other ways. Isn’t it?</blockquote><p>2. Well, I have to transfer money at any cost. So, I wrote down her account number on a tissue paper near me.</p><p>3. Although my handwriting was large enough to read, the screen where I was keying in those 12 digit numbers was too small, including the font size.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EYCX5t14KsUdu7-LOPzf8g.png" /><figcaption>Not too good experience. At least INCREASE the font size guys!!! 😤</figcaption></figure><p>I will cover the rest of the transfer process in Part 2. For now, let’s see what we can improve here.</p><h3>Solutions</h3><ol><li>Switching apps logged me out from CB App. I do not know the reason why it is done this way but…</li></ol><blockquote>Team! you need to let users switch between apps.</blockquote><p>2. Chunking the numbers like this. Check out the improved design below.</p><blockquote><strong>0086 6001 0018 7275</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1010/1*pPKTuejjFb6MvsgdcNI7eQ.png" /><figcaption>Chunking 12 digit account number</figcaption></figure><p><a href="https://xd.adobe.com/view/1df64fc9-dce6-48a8-4dce-b90d3d900597-6bc7/">CB Bank Mobile App - Redesign Part 3</a></p><p>*I do not work for CB Bank. This is just a volunteer contribution for CB Mobile App users. I wish CB Mobile App team read this and improve their app way better than the design suggested here.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fe5e2e4418c0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Growth Strategy for Vedic Chart App]]></title>
            <link>https://medium.com/@dhanmoti/growth-strategy-for-vedic-chart-app-32aaa0ca0ab?source=rss-34d60b1d76d1------2</link>
            <guid isPermaLink="false">https://medium.com/p/32aaa0ca0ab</guid>
            <category><![CDATA[mobile]]></category>
            <category><![CDATA[app-store]]></category>
            <category><![CDATA[in-app-purchase]]></category>
            <category><![CDATA[itunes-connect]]></category>
            <category><![CDATA[sales]]></category>
            <dc:creator><![CDATA[Dhan Moti]]></dc:creator>
            <pubDate>Wed, 20 Jun 2018 12:41:13 GMT</pubDate>
            <atom:updated>2018-06-20T12:41:13.523Z</atom:updated>
            <content:encoded><![CDATA[<p>Recent version of <a href="https://vedicchartapp.com/">Vedic Chart App</a> uses some of medium In-App Purchases behaviour and has made <strong>62% increase</strong>. Next version will be infusion of both <a href="https://medium.com/p/32aaa0ca0ab/edit">Medium</a> and <a href="https://www.uber.com/en-SG/">Uber</a> to increase user base without ads.</p><h3>62% increase in In-App Purchases for last 30 days.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/539/1*_8NrCuAM0lzq6srUwHkLzA.png" /><figcaption>Analytic from iTunesconnect / appstoreconnect</figcaption></figure><blockquote>New Strategy Coming soon…on Medium. Stay tuned!!!!</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SuJoZ2UMDQ1fdaM9lB7Geg.jpeg" /><figcaption>To increase users/download on its own</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=32aaa0ca0ab" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>