<?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 Atmika Pai on Medium]]></title>
        <description><![CDATA[Stories by Atmika Pai on Medium]]></description>
        <link>https://medium.com/@atmikapai?source=rss-63b28313683a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*QuxGfRCuWlFMw4ALk4vNfg.png</url>
            <title>Stories by Atmika Pai on Medium</title>
            <link>https://medium.com/@atmikapai?source=rss-63b28313683a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 16:45:12 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@atmikapai/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[On Daniel Dennett’s Intentional Stance and the Question of True Belief]]></title>
            <link>https://medium.com/@atmikapai/dennetts-intentional-stance-and-the-question-of-true-belief-f8d564bebd43?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/f8d564bebd43</guid>
            <category><![CDATA[philosophy]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Thu, 10 Sep 2026 14:40:40 GMT</pubDate>
            <atom:updated>2026-09-10T14:41:46.575Z</atom:updated>
            <content:encoded><![CDATA[<p>Belief attribution is a contentious subject in philosophy. Physical realists view belief as a physical event in the brain, while interpretationists acknowledge that belief attribution is relativistic. Daniel Dennett concedes that while belief is an objective phenomenon, it can be confirmed by a certain kind of interpretation, that is to say a certain predictive strategy. If the predictive strategy is successful in forecasting the system’s behavior, then the system does indeed hold beliefs. Dennett explicates this in his thesis in <em>True Believers</em> — a system whose behavior is well predicted by the intentional stance (Dennett’s predictive strategy), then the system is an intentional system, and we can conclude that the system is a true believer. This is a melange of the two viewpoints. In this paper, I will elaborate on Dennett’s thesis by outlining the key concepts, then I will evaluate his claims with examples.</p><p>Dennett believes that a true believer is an intentional system if its outcomes and behaviors can be predicted via the intentional strategy. I will define the key terms, followed by an example. According to Dennett, a system is any object. He purposely selects a neutral term, as he wants to avoid revealing any connotations of what the object can or cannot do. A strategy or stance inculcates a set of rules or a point of view that one can undertake to understand and predict the behavior of a system. What follows next is the intentional strategy, which is one of the three predictive strategies outlined by Dennett. The intentional strategy involves treating the system whose behavior we want to forecast as a rational agent with beliefs, desires, and mental states exhibiting intentionality. A rational agent will always act in a manner that is aligned with its goals. Intentionality refers to the representational character of the mind. Another way to look at intentionality is to understand a system as being oriented towards something. Next, Dennett urges us to use practical reasoning to ascertain these propositional attitudes keeping the system’s environment and context in mind. He probes us to question — “what ought the system to believe […] and desire considering its place in the world and its purpose?” (Dennett, 343) Some might argue that this question produces an exhaustive list of answers of which some are not even discernible unless we specifically interrogate the system. But, Dennett advocates to consider only the beliefs and desires that are relevant to the system’s interests and that the system’s experience has made available, and assume that the system holds no contradictory beliefs and desires. Finally, upon applying the strategy, we should get a prediction of the system’s behavior or outcome. If the predictions via the intentional strategy are successful, then the system is an intentional system, and therefore, a true believer.</p><p>Let me clarify Dennett’s intentional systems through an example: If A is fond of macaroons, and I offer A a macaroon, then we can predict with certainty that A will accept the macaroon. This is because we apply the intentional strategy by assuming A is a rational agent. We know A desires macaroons because it has a predilection for them. As A believes macaroons to be good for it, we surmise that it ought to accept it. This example illustrates that A is a true believer, as it is an intentional system whose behavior can be predicted by applying the intentional strategy. I chose this example because it illustrates the act of gift-giving without defining A as a human and the macaroons as a gift. I hope this demonstrates that not only do we use the intentional stance to shape our expectations of systems daily but also how well-suited the strategy is for humans. According to Dennett, the strategy works effectively because there is some pattern to human action, namely our propensity to have beliefs and desires. Therefore, the intentional strategy helps abstract away from the inner mechanics of the mind and the physical composition of bodies and instead hone in on these patterns. The abstraction also enables the intentional strategy to predict the outcomes for complex systems beyond humans, and this is also why Dennett doesn’t conflate having a good strategy with knowing why it works.</p><p>Some might not be convinced by the intentional strategy to attribute beliefs. They might argue that A <em>acts</em> like it desires macaroons while a human who loves macaroons <em>really</em> desires the dessert. They would concede that while we did accurately predict A’s behavior, that doesn’t necessitate that A is a true believer and an intentional system. Dennett rejects this claim because he believes there is no distinction between really desiring vs mimicking desire. Let me illustrate this through a counter-example: a brick is situated in a building, and we must formulate the brick’s behavior and outcomes. We could apply the intentional strategy by viewing the brick as a rational agent that desires to keep the construction of the building sound. Alternatively, we could use Dennett’s design strategy, which appeals to the design of an object to predict its behavior under various conditions, or Dennett’s physical strategy, which uses the laws of physics to predict the outcome of a system for any input. All three examples would conclude that the brick ought to maintain its position in the building’s construction. While it is up to us to use any of the three stances, Dennett would caution that facts of the strategy’s success and predictive power are objective. Thus, it would be futile to use the intentional strategy, as it does not shed light on the brick’s outcomes any more than the design and physical strategies do. Therefore, the brick is not a true believer nor an intentional system, following Dennett’s thesis. This example illustrates that the intentional strategy is specifically geared towards isolating intentional systems, because the strategy is only admissible when it has predictive power that couldn’t be acquired through the other strategies. Thus, if the intentional strategy is effective in describing the behavior of a system, then the system is a true believer, and it does indeed have the beliefs and desires that we attribute to it when applying the intentional stance.</p><p>Let us revisit the example of A and its predilection for macarons. If A were to accept the macaroon, it should be clear that A does indeed desire macaroons, and A is a true believer. A’s design is not specified in the example above, but say it was revealed to us that A is designed to be rational and desire what it ought to want. Then, we could apply the design stance and surmise that A is designed to fulfill its desires, and it desires macarons, therefore, it ought to accept the macaroon. At this point, you may raise two points: (1) A’s design is similar to the assumptions Dennett outlines in the intentional stance, and (2) if we could predict A’s outcome using the design stance instead of the intentional stance, then A is no longer a true believer. To the first claim, I would argue that the intentional stance is a subspecies of the design stance because evolution did design us to be rational, desire what we ought to want, and believe what we ought to think. Like the design stance, the intentional stance abstracts away from the actual physical implementation and takes on an instrumentalist viewpoint. I think Dennett would agree with me on this. As for the second claim, I would say that if A were an intentional system, then the intentional strategy would likely have higher predictive value in other behavioral predicaments of A, therefore, those analyses would indicate that A is a true believer.</p><p>Lastly, I want to consider the possibility that A refuses the macaroon. This raises questions of the validity of our predictions as well as ascertaining A as a true believer. Dennett would respond by asking us to re-examine and revise our ideal of perfect rationality. He doesn’t expand on the extent to which we should revise our model and the impacts of this on our prediction, which is a limitation of the strategy, however, I believe it is also outside the scope of <em>True Believers. </em>Another point to explore is the role of opinions in influencing our behavior and how opinions fit into Dennett’s schema of intentional systems. For example, how would our predictions of A change if A held the opinion that macarons are the least healthiest form of dessert. It is outside the scope of this paper, but it would be a worthy investigation.</p><h3><em>Works Cited</em></h3><p>Dennett, Daniel Clement. “True Believers: The Intentional Strategy and Why It Works.” The</p><p>intentional stance. MIT press, 1989.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f8d564bebd43" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Rules & Principles in Grice’s Theory of Conversation]]></title>
            <link>https://medium.com/@atmikapai/rules-principles-in-grices-theory-of-conversation-0d719c5fc668?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/0d719c5fc668</guid>
            <category><![CDATA[philosophy]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Thu, 10 Sep 2026 14:32:35 GMT</pubDate>
            <atom:updated>2026-09-10T14:32:35.530Z</atom:updated>
            <content:encoded><![CDATA[<p>We have all been in a situation where we meant to say one thing but mean another. In philosophical jargon, this is equivalent to the linguistic meaning or literal meaning of a sentence diverging from the meaning that one aims to convey with the sentence. Yet, often enough, addresses in conversation are able to grasp the particular intended meaning by considering the context or perhaps assuming that the speaker is rational and aims to deliver an informative, useful comment relevant to the topic of discussion at hand. Logicians who attempt to construct a systematic theory of meaning find these situations challenging, because there isn’t a one-to-one correspondence between linguistic form and utterance meaning. In <em>Logic and Conversation, </em>H.P.<em> </em>Grice shows how to leverage our general intuition that speakers are rational to explain how the speaker’s intended meaning in a context is transmitted beyond the literal meaning of the sentences involved. His work proves useful to logicians as he demonstrates that context sensitivity and indirect speech acts are not intractable obstacles to a systematic theory of meaning of natural language. In this paper, I explain Grice’s Cooperative Principle (CP) and the four maxims that articulate our general rationality and intuition when in conversation. Then, I will elaborate on his notion of conversational implicatures and flouting through examples. Finally, I evaluate the applicability of Grice’s theories to real-life contexts.</p><p>Grice observes there are assumptions and expectations in ordinary language that enable interlocutors to grasp a speaker’s intended meaning that goes beyond the literal meaning of what the speaker says. He believes that these expectations are grounded in rationality and stem from intuition. He posits the Cooperative Principle to describe these conversational norms or etiquettes as follows: “Make your conversational contribution such as is required at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” (Logic and Conversation, 45) What he means by this is that when one says or hears a sentence in conversation, one assumes that it will be informative, true and provable, relevant to the topic of discussion, and coherent. These expectations can be categorized into four maxims — Quantity, Quality, Relation, and Manner. To reiterate, quantity mandates that the utterance is informative but no more than required, and quality implies that the contribution of the remark is true and provable by adequate evidence. Relation entails that the statement is relevant to the topic of discussion, and finally, manner signifies that the speaker is coherent and intelligible in his remark.</p><p>Let us look at an example to realize the applicability of Grice’s CP: a renter is visiting apartment showings, and he asks the landlord, ‘Is the apartment available for rent?’. The landlord responds, ‘Yes, the lease starts June 1, 2022.’ This is an example when the speaker obeys the maxims of CP. He believes that the lease starts June 1, 2022, that the renter does not know this, and thinks this piece of information is relevant to the conversation at hand. While the landlord could have simply said yes and achieved the Quality maxim governing succinctness, the lease start date is relevant to the question. Grice allows for such details because they adhere to the maxims of CP. Thus, this example demonstrates how the maxims of CP are at play when CP governs the conversation.</p><p>A speaker may often utter information adjacent to the conversation and expect the listener to infer or pick up on the relevance of his remark. While this violates the Quantity and Relation maxim, the listener is in a position to glean information by cooperating and engaging in the conversation. Grice observes that this tension between what the speaker says and the maxims of CP may still allow the listener to infer information. He terms this exploitation of maxims as flouting and the inferred pieces of information as conversational implicature. One thing to note is that, in these scenarios, CP still governs the conversation because the maxims have been exploited rather than violated. In other words, when a speaker’s utterance doesn’t conform to the maxims in CP, then Grice believes that an appropriate meaning is there to be inferred in virtue of the interlocutors mutually prescribing to CP. Thus, the listener is in a position to infer implicatures, because he can leverage the assumptions stipulated in CP. While Grice concedes that CP governing the conversation is taken for granted, he argues that the speaker can leverage background assumptions about rationality and cooperation to minimize the discrepancy in the conversation and reconcile with the speaker, which is ultimately a benefit for the interlocutors involved. This is the shortest course of action for the listener to grasp the speaker’s particular intended meaning without straying too far from the conversation. For Grice, flouting a maxim and generating an implicature also justifies why interlocutors take the effort to interpret and understand an indirect speech act, an action that would otherwise be fruitless.</p><p>Continuing with the renter’s example, let us consider the situation where the landlord responds to the renter’s inquiry by saying, “Some kitchen work is remaining.” Since the landlord is cooperative under CP, the renter’s task is to figure out how to construe this remark so that it is a cooperative contribution to the conversation. The renter might reason, for instance, that the apartment in question is being renovated, and the kitchen still requires some work out of all the rooms in the apartment. So, the renter might surmise that the apartment isn’t ready for leasing. The idea that the speaker believes that the apartment is being renovated enables the listener to construe the remark as cooperative after all. The implicature is that the kitchen in the apartment still needs some work. Thus, the renter can surmise that the apartment isn’t ready for rent just yet.</p><p>One thing to note is that in the scenario above, the Relevant maxim has been flouted, and so an implicature was generated. If the landlord blatantly lies or inadvertently gives a long-winded response to the renter’s question, then the speaker is simply violating the maxims of CP. For example, if the landlord has already decided to lease the apartment to another tenant and answers discontent to the current renter’s question, then the maxims of CP have simply been violated and CP no longer governs the conversation. Some might oppose Grice’s theory by probing how does the renter know that the landlord has violated their mutual agreement to adhere to the maxims of CP? If the renter is unaware of this, then he will continue to try to be cooperative with the landlord to his own detriment. Grice would likely respond by saying that it is important that all the interlocutors mutually adhere to CP, as only then do the maxims of CP apply. It doesn’t matter if one is violating and the others are in agreement with the maxims because CP does not apply then.</p><p>One might also reason that it is not necessarily beneficial for interlocutors to disclose all the information that they know is relevant to the conversation. For example, in the case above, the relationship between a renter and landlord hinges on asymmetric information, which is economic jargon for the situation in which one party in an economic relationship or transaction is in ownership of more information than the other party. It is nearly impossible for a renter to know all the apartments and their respective rents in the market, and the landlord can take advantage of his lack of knowledge to upcharge him. This would be an act where the speaker is not cooperating with the listeners. The relationship between the interlocutors would likely shed light on the misgivings in the conversation. I wonder if Grice would be open to considering the relationship of the interlocutors to make his CP theory more robust.</p><p>Some might also argue that it is unclear whether the four maxims are rules or principles. If the maxims are rules, then they are either in effect or not. This leaves no room for gray areas, and there are more situations in which maxims are flouted than maxims obeyed. Alternatively, if the maxims are principal, then they can be achieved to varying degrees. For example, the Manner maxim presupposes that the speaker can speak fairly and succinctly when in conversation, however, a speaker may not be proficient in the language that the conversation is taking place in and do his best to converse with his limited vocabulary. Would this be a violation of Grice’s CP? While Grice may dismiss these edge cases, they are compelling scenarios that arise in everyday situations, especially in our globalized world. In a similar vein, CP aptly addresses the context-sensitivity of natural language, however, it is not culturally abiding. For example, in India, it is considered extremely rude if you speak succinctly and give simple, brusque answers when you’re talking to your elderly. This is because it is believed that one should give thoughtful, insightful, and engaging responses to the elderly to show that they respect them. According to Grice, conversations like these would violate the Manner maxim. I wonder how he would respond to cultural differences.</p><p><strong>Works Cited</strong></p><p>Grice, H. P. (2002). Logic and conversation. In D. J. Levitin (Ed.), <em>Foundations of cognitive psychology: Core readings</em> (pp. 719–732). MIT Press.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0d719c5fc668" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[NYC Eats: New Digital Interfaces for Maps & Chatbots]]></title>
            <link>https://medium.com/between-the-graticules/nyc-eats-new-digital-interfaces-for-maps-chatbots-5e562694d671?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/5e562694d671</guid>
            <category><![CDATA[maps]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Sun, 15 Feb 2026 23:24:11 GMT</pubDate>
            <atom:updated>2026-02-15T23:24:11.747Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_uDVGjbX0waq7LTct5eWcA.png" /></figure><p>NYC Eats is an agentic restaurant concierge MVP that reimagines how New Yorkers discover restaurants. Forget scrolling through pins on a map and parsing star ratings i.e. the dominant paradigm since Google Maps. NYC Eats features Remi, a conversational agent who guides users through the messy, multi-variable way people actually make dining decisions.</p><p>A typical query might be: “My friend’s in Chelsea, I’m still in Midtown at work. We’re willing to transit 15 minutes max. Find us happy hour spots somewhere between us.”</p><p>Or: “Cozy date night. My girlfriend is gluten-free. Find me eclectic, hole-in-the-wall spots.”</p><p>These are real-life scenarios, with travel constraints and subjective restaurant attributes, that conventional apps like Google Maps, Beli, or Yelp dismiss entirely.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HyQJruWf5F9gCR9LX-kBLg.png" /></figure><p>Google added some Gemini chat capabilities into Maps, when they made a massive push to integrate AI across their entire product suite. But, when you hastily patch a chatbot onto your product, the results are unimaginative and has little to do with improving user experience. It’s also a quick, cheap way of “AI-fying” your current product, what I think Scott Belsky, founder of Behance and now leading A24 labs, would call “skeumorphic” design. This describes the UI principle of new tech mimicking the design of old tech to ingratiate users in the first-mile experience of new tech. With the new tech being the ability to prompt and chat with LLMs to decipher complex queries, this would translate to having a chatbot sitting next to the map without any foresight of actual integration.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4F_D9XKeD0RHGemlcdkHVg.png" /><figcaption>Prompt: “Japanese restaurants with 4 rating or higher within 10 min transit of me”</figcaption></figure><p>Now, NYC Eats imagines what a more daring, opinionated, and unconventional integration of Gemini with Google Maps could look like. This means, exploring fresher interfaces: the chat is directly embedded in the map, which in-turn dynamically creates isochrones, resizes, and zooms, giving the user the feeling that they’re talking to the map itself.</p><p>NYC Eats also strays from a bland, faceless chatbot, instead featuring a overzealous Remi, inspired by Remy from Ratatouille and New York native epicurean Anthony Bourdain.</p><p><strong>The Inspiration &amp; Motivation?</strong></p><p>NYC Eats is inspired by two schools of thought. The first is technological: apps like Dennis Crowley’s Foursquare and Dodgeball treated cities as living, discoverable spaces. Google Maps acquired Dodgeball and basically shut down the project. I’m trying to bring back the energy of Dodgeball and Foursquare with NYC Eats.</p><p>The second is Guy Debord. In the 1960s in Paris, Debord had an epiphany: urban life has become too routinized, too dictated by commerce and efficiency, and people weren’t actually experiencing Paris. To circumvent that, he derived the concept, “Dérive”, getting lost on purpose, letting mood guide where you go instead of a fixed route. Google Maps is the opposite — optimized for efficiency: get from A to B, sort by rating, done.</p><p>Google Maps is the opposite of that. It’s optimized for efficiency: get from A to B, sort by rating, done. And because digital tools increasingly structure how people navigate physical space, those same categories — price, distance, stars — become the only lens through which people find restaurants.</p><p>And so, NYC Eats tries to reintroduce subjectivity and ambiguity; asking Remi for “something eclectic and hole-in-the-wall” is closer to a digital dérive than a database query.</p><p><strong>The So What? &amp; Why?</strong></p><p>I think this personalization or curated, location-aware user experience is where spatial apps are headed, not just Google Maps, and I’m exploring the new visual interfaces and mediums that come with this shift. I hope NYC Eats is a compelling provocation of what the future heralds, inspiring urban, creative technologists.</p><p>Finally, I believe cities should have a slew of homemade urban apps that are better at local recommendations precisely because they’re built by native residents. This is part of my broader open-source philosophy: that the best city tools should come from the city itself. Vibe-coding tools and agent frameworks have now democratized software development to the point where a single person can build what once required a team à la NYC Eats. As an Urban Tech grad student at Cornell Tech, I’m positioned at the intersection of these tools and want to show people what can be built with them.</p><p><strong>Tech stack:</strong> Gemini as reasoning agent, multi-tool orchestration via LangChain, geospatial toolkit for dynamic map rendering, and RAG-based semantic search for curated, contextual recommendations.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5e562694d671" width="1" height="1" alt=""><hr><p><a href="https://medium.com/between-the-graticules/nyc-eats-new-digital-interfaces-for-maps-chatbots-5e562694d671">NYC Eats: New Digital Interfaces for Maps &amp; Chatbots</a> was originally published in <a href="https://medium.com/between-the-graticules">Between the Graticules</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tutorial: ArcGIS Maps SDKs for Unity]]></title>
            <link>https://medium.com/between-the-graticules/tutorial-arcgis-maps-sdks-for-unity-678ea28ebc73?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/678ea28ebc73</guid>
            <category><![CDATA[mapping]]></category>
            <category><![CDATA[unity]]></category>
            <category><![CDATA[esri]]></category>
            <category><![CDATA[3d]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Mon, 10 Nov 2025 15:18:27 GMT</pubDate>
            <atom:updated>2025-11-10T15:27:24.309Z</atom:updated>
            <content:encoded><![CDATA[<h4>A walkthrough of how to set up Esri ArcGIS Maps SDKs for Unity with sample project and video demo</h4><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FALznzZ44UMs%3Fstart%3D40%26feature%3Doembed%26start%3D40&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DALznzZ44UMs&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FALznzZ44UMs%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/cfdecbe9fa48b749315d59b333094709/href">https://medium.com/media/cfdecbe9fa48b749315d59b333094709/href</a></iframe><p>The theme of Day 6 of <a href="https://30daymapchallenge.com/">#30DayMapChallenge 2025</a> was to go 3D:</p><blockquote>Map beyond 2D. Visualize data using 3D models, extrusions (building heights), depth, time (as a dimension), or an unconventional multivariate approach.</blockquote><p>As a GIS technologist interested in Augmented and Virtual Reality (AR/VR) experiences, I was curious what kind of AR/VR application and tools were at urban planners, architects, and designers’ disposal.</p><h4>History of 3D Applications</h4><p><a href="https://unity.com/">Unity</a> is primarily a game engine, helping developers create 2D and 3D games. It began as a game engine in 2005 with the goal of democratizing game development for Mac users, a less flexible and malleable operating system than its counterpart, Windows. Since then, it has expanded to support interactive 3D content for industries like film, automotive, and AR/VR applications. If you are a gamer, you may have already come across it with major game studios like Ubisoft and Electronic Arts, as well as companies like Niantic and Meta now deploying it.</p><p><a href="https://gravitysketch.com/">Gravity Sketch</a> is a 3D creation platform, helping users sketch and collaborate in virtual reality (VR) environments. The platform is used by professionals and educators in industrial design, entertainment, and AR development. Here’s a couple of my rudimentary designs in Gravity Sketch, part of Cornell Tech’s Augmented and Virtual Reality coursework.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dp5n3Bz8bH78QMAWbv6m7w.png" /><figcaption>Two dogs hanging out</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*uaEwPf_HscgChTvm9ArGJg.png" /><figcaption>Rudimentary motorcyle</figcaption></figure><p>Think of ArcGIS Maps SDKs for Unity as a combination of these two in that you develop in Unity to give a Gravity Sketch-like feel. I would say, the primary target demographic is architects, urban planners, and designers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/995/0*XYyJgKgE2hV48vKz.png" /><figcaption>Sample Viewer Project running in Unity (<a href="https://developers.arcgis.com/unity/sample-code/">Documentation</a>)</figcaption></figure><p>Let’s take a look how to set up the <a href="https://github.com/Esri/arcgis-maps-sdk-unity-samples#instructions">Sample Viewer,</a> which hosts <a href="https://github.com/Esri/arcgis-maps-sdk-unity-samples?tab=readme-ov-file#features">15 features for spatial analysis and design</a>: the third-person controller lets you navigate cities as a character, while the viewshed analysis calculates sight lines from any point — I tested this from the Empire State Building’s observation deck.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lb3EmD1nr7Kq3HsaQF6hvg.png" /><figcaption>15 features available in Sample Viewer project (<a href="https://github.com/Esri/arcgis-maps-sdk-unity-samples?tab=readme-ov-file#features">Github</a>)</figcaption></figure><h3>ArcGIS Maps SDKs for Unity Set-up Instructions</h3><p>While Esri provides documentation on how to set up the Sample Viewer project, I got list in the weeds. So, I have listed out more thorough instructions for Unity neophytes like myself:</p><ol><li>Clone the Sample Viewer <a href="https://github.com/Esri/arcgis-maps-sdk-unity-samples">Github</a> repo.</li></ol><p>Note: There will be three projects within this one file: 1) sample_project and 2) xr_sample_project.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LXNpA-1CtPiL_FITjxmXOw.png" /><figcaption>Three Unity files bundled into one when you clone Github Repo</figcaption></figure><p>2. Assuming you have already installed Unity, navigate to Unity Hub. Click Add &gt; Add project from disk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qM5wVsidttrgqQtumVeXrQ.png" /><figcaption>Opening cloned Github repo in Unity</figcaption></figure><p>Navigate to your cloned Github repo, and open sample_project in Unity. It will ask you to download the dependencies and correct Unity version, namely Unity 2022.3.62f2 . (Downloads will take about 30–45 minutes.)</p><p>3. Open the project once all the dependencies are downloaded. You should land on this configuration in Unity:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JoXJQYG2i1KAcY8-ZQ3o9w.png" /><figcaption>Landing page once you open the cloned Github repo in Unity</figcaption></figure><p>You can see the Sample Viewer in the Projects console on the left.</p><p>4. Download and install the ArcGIS SDKs plugin from this <a href="https://developers.arcgis.com/unity/install-and-set-up/#download-from-the-downloads-page">link</a>.</p><p>5. After downloading, in the top menu in Unity, expand Window &gt; Package Manager.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ioz5MKZEbc0a4QGMFBAi8A.png" /><figcaption>Opening Package Manager in Unity to add ArcGIS Maps SDKs plugin into project</figcaption></figure><p>Next, Click the + icon in the Package Manager window.</p><p>Select Add package from tarball from the list and open a file browser.</p><p>Navigate to the folder where you saved your tarball file.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9MacrjF_Mqug51bp2JYATg.png" /><figcaption>Adding ArcGIS Maps SDKs plugin from tarball</figcaption></figure><p>Once the plugin has been downloaded, you will see the following page in Unity’s Package Manager:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1eX82XOX1upE3lzTvUiOBA.png" /><figcaption>Confirmation that ArcGIS Maps SDK has been downloaded</figcaption></figure><p>6. Next, navigate to Samples within ArcGIS Maps SDK in Package Manager, and import the samples.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*buDKiWVSKfKew88ucLRnkg.png" /></figure><p>7. Next, ou will have to next get an API key, and here are the <a href="https://developers.arcgis.com/unity/get-started/#3-get-an-api-key">instructions</a> for that. I would recommend setting up a free ArcGIS Location Platform account for it.</p><p>8. To add the API key to your project, got to the top menu in Unity, and expand Edit &gt; Project Settings.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HX4_uCwqaIS3SOOu3otMPA.png" /><figcaption>Navigat to Edit &gt; Project Settings to add API key</figcaption></figure><p>Add the API key here. I found this to be the easiest way to embed our API key into the project.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KFL9xTan7xpJ-q_XdjRgdQ.png" /><figcaption>Adding API key to Unity project</figcaption></figure><p>And you’re all set!</p><p>To get the Sample Viewer running, navigate to SampleViewer folder, and click on the Sample Viewer script. Hit play in the top menu in Unity, and you’re ready to play.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wM5agJOpKg64Zph2Kcw2QQ.png" /></figure><p>Take a look at the video at the top of this article that goes through all 15 features in the Sample Viewer project so you can get a sense of what the future of urban planning in AR/VR looks like!!</p><p>Also, check out the other maps my Cornell Tech classmates and I are putting together for #30DayMapChallenge 2025:</p><p><a href="https://bsky.app/profile/urban-data.bsky.social">Urban Data (@urban-data.bsky.social)</a></p><p>Happy mapping!</p><h4>Resources</h4><ul><li><a href="https://developers.arcgis.com/documentation/security-and-authentication/get-started/">Get started | Documentation | Esri Developer</a></li><li><a href="https://github.com/Esri/arcgis-maps-sdk-unity-samples">GitHub - Esri/arcgis-maps-sdk-unity-samples: Sample code for the ArcGIS Maps SDK for Unity.</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=678ea28ebc73" width="1" height="1" alt=""><hr><p><a href="https://medium.com/between-the-graticules/tutorial-arcgis-maps-sdks-for-unity-678ea28ebc73">Tutorial: ArcGIS Maps SDKs for Unity</a> was originally published in <a href="https://medium.com/between-the-graticules">Between the Graticules</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My Favorite Cartographers of All Time]]></title>
            <link>https://medium.com/between-the-graticules/notes-on-esris-cartography-mooc-f0c4040f1d01?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/f0c4040f1d01</guid>
            <category><![CDATA[gis]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[esri]]></category>
            <category><![CDATA[maps]]></category>
            <category><![CDATA[spatial-analysis]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Tue, 14 Oct 2025 14:14:21 GMT</pubDate>
            <atom:updated>2026-07-16T13:11:01.846Z</atom:updated>
            <content:encoded><![CDATA[<p>Every year, Esri runs a series of free <a href="https://www.esri.com/training/mooc/">massive open online courses</a> (MOOCs), training users on how to apply geospatial thinking and create data-driven insights. I had the opportunity to attend the <a href="https://www.esri.com/training/catalog/57660f19bb54adb30c9454b0/going-places-with-spatial-analysis/"><em>Going Places with Spatial Analysis</em></a> and <a href="https://www.esri.com/training/catalog/596e584bb826875993ba4ebf/cartography/"><em>Cartography</em></a> MOOC. I wanted to share my notes on the latter course, as I got to learn about best practices in geospatial modeling and the latest ArcGIS Pro 2D and 3D Map design tools from experienced professionals.</p><p>Esri’s <a href="https://www.esri.com/arcgis-blog/author/kenfield">Kenneth Field</a> runs this MOOC with a cartographer that has gained celebrity status, <a href="https://www.esri.com/arcgis-blog/author/j_nelson">John Nelson</a>. This motley crew is accompanied by three other rotating guests — <a href="https://www.esri.com/arcgis-blog/author/epunt">Edie Punt</a>, <a href="https://www.esri.com/arcgis-blog/author/wesjones">Wes Jones</a>, <a href="https://www.linkedin.com/in/nathan-shephard-0622912a/">Nathan Shepherd</a>. Each week begins with a 20-minute presentation on the fundamentals, where our hosts walk users through best practices, Cartographer and Map of the Week, Cartofails, or otherwise known as pet peeves in map-making. Next, there’s short, practical exercises in ArcGIS Pro with a discussion forum for learners to share their queries, insights, and feedback.</p><p>I was so excited to hear about famous cartographers and maps in the MOOC that I wanted to make a post that I could always refer back to. Below, the mapmakers are presented in chronological order, so you can trace the history of mapmaking across time and region.</p><h3>An Assortment of Cartographers Over Time</h3><p><strong><em>Eratosthenes of Cyrene (194 BC) </em></strong>was an Ancient Greek polymath with interests in poetry, astronomy, and of course, geography. A true savant, he became the chief librarian at the Library of Alexiandra. He was also the first to calculate the Earth’s circumference and its axial tilt with commendable accuracy. He also created the first global projection of the world, incorporating parallels and meridians. Retrospectively, he is deemed as the Father of Geography.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/245/0*NEtHYaqNfuOi211e" /><figcaption>Source: <a href="http://www.geo.hunter.cuny.edu/~jochen/gtech201/lectures/lec6concepts/Datums/Determining%20the%20earths%20size.htm">Hunter College</a></figcaption></figure><p><strong><em>Charles Minard (1781–1870), </em></strong>a French civil engineer, is renowned for his illustration of Napoleon’s failed invasion of Russia in 1812. A Sankey diagram before it’s time, the illustration below illustrates the size of Napoleon’s army during their advance and retreat through the thick black and beige lines, respectively. Within two dimensions, the map captures six types of data: lat/long, distance and direction travelled, number of Napoleon’s troops, and temperature. One quickly realizes that the minutae within this map comes moreso from Minard’s compassion for the travails and sacrifices of the soldiers rather than statistical ambition. Read more about this venerable map <a href="https://www.esri.com/content/dam/esrisites/en-us/esri-press/book-pages/sample-page/mapping-time-illustrated-minards-map-napoleons-russian-campaign-1812.pdf">here</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ud3RYczVDM1SuhbOw7n1wA.png" /><figcaption>Source: <a href="https://en.wikipedia.org/wiki/Charles_Joseph_Minard#/media/File:Minard.png">Wikipedia</a></figcaption></figure><p>Everyone’s encountered <strong><em>John Snow’s (1813–1858) </em></strong>illustration of London’s deadly cholera epidemic<strong><em>,</em></strong> but did you know that Snow’s strategic depiction of water pumps in the map helps trace it back to the one responsible for the outbreak. This kind of a representation is known as a Voronoi diagram or Thiessen polygons. It divides a map into regions based on proximity to specific points. Each polygon represents an area of influence around its point — meaning that any location within that polygon is closer to its point than to any other. It’s similar in spirit to an isochrone analysis, which shows areas reachable within a certain time or distance usually via public transit station or stop.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/966/1*1JD3zFfukGpH1wXp6BSEiA.png" /><figcaption>Source: <a href="https://education.nationalgeographic.org/resource/mapping-a-london-epidemic/">National Geography</a></figcaption></figure><p><strong><em>Étienne-Jules Marey (1830–1904), </em></strong>a French scientist, physiologist, chronophotographer by profession, has created one of my favorite maps (seen below). As a transit enthusiast and oppidan like myself, this maps retains all the necessary commuter information without inundating the user as Google Maps is wont to do.</p><p>The chart plots distance traveled against the time taken. The distance is represented by the station names along the Y axis and the arrival or departure time at the corresponding station along the X axis. The station names are displayed in proportion to the distance between them. This kind of chart gained notoriety, gaining the monicker Marey chart after the creator himself! Replacing boring time tables with Marey charts may improve information literacy, an essential skill in the information and digital age. Plus, they are way more visually appealing!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GjNX8mJha7qP0xZJaIFE8w.png" /><figcaption>Source: <a href="https://www.3cs.ch/analyze-work-flow-with-marey-chart/">The View of Flow Management blog</a></figcaption></figure><p><strong><em>Erwin Raisz (1893–1963), </em></strong>acclaimed for his physiographic maps of landform, was the first to author a cartography textbook. Raisz’s choice of serif font to the watercolor effect of the colors is reminiscent of an antiquated time. Aren’t you immediately transported back to your geography classroom?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8SkVXXqsHdACc_GLcoDejQ.png" /><figcaption>Source: <a href="https://mapdesign.icaci.org/2014/04/mapcarte-117365-atlas-of-global-geography-by-erwin-raisz-1944/">ICA Commission on Map Design WordPress</a></figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*i9pz6u9DX0PZvn4P.png" /><figcaption>Source: <a href="https://mapdesign.icaci.org/2014/04/mapcarte-117365-atlas-of-global-geography-by-erwin-raisz-1944/">ICA Commission on Map Design WordPress</a></figcaption></figure><p><strong><em>Jacques bertin (1918–2010) </em></strong>was a French cartographer and theorist, known for his seminal work, <a href="https://www.esri.com/en-us/esri-press/browse/semiology-of-graphics-diagrams-networks-maps"><em>Semiology of Graphics</em></a>. I’ve eyed this mammoth book numerous times at various Esri conferences. One day, it’ll be in my collection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/318/1*7AsuTjfEg2SmA9PLcUKltQ.png" /><figcaption>Semiology is the study of signs and symboles and their use or interpretation. Source: <a href="https://www.esri.com/en-us/esri-press/browse/semiology-of-graphics-diagrams-networks-maps">Esri Press</a></figcaption></figure><p><strong><em>Heinrich Berann (1915–1999), </em></strong>an Austrian painter and cartographer, melds the two fields to create virtuosic panoramic maps. He gained notoriety with tremendous run of over 100 maps, working with the Olympic games, U.S. National Park Services, and National Geographic Magazine.</p><p>My favorites in his collection are the Yellowstone and Yosemite National Park (former seen below). The saturated colors with the addition of the clouds makes my jaw drop every time. Also, I appreciate that he avoids photorealistic, hyperrealistic representations, eliminating pin-needle accurate depictions of heights, sizes, and dimensions. In other words, Berann obfuscates the challenges that come with capturing a landscape on your phone where the depth is all distorted by developing his own style of “mountainscape painting”. Here, he plays with depth and height in interesting ways, transcending photorealistic depictions, instead capturing what the eye sees and how it processes the scene.</p><p>I’d say Berann is successor to Raisz’s maps. I would have loved to see Berann’s depictions of Mount Fuji. The Himalayas may will suffice! Fun fact: mountain climbers gathered over 100 photos of the Himalaya mountain range for Berann, who then then spent 600 hours to paint the mountain-scape.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BiVB100x0DJ6VPp8orICVA.png" /><figcaption>Source: <a href="https://en.wikipedia.org/wiki/Heinrich_C._Berann">Heinrich Berann’s Wiki</a></figcaption></figure><p><strong><em>Andy Woodruff </em></strong>is a contemporary Bostonion cartographer with nearly 15+ years experience in making interactive web maps. I was surprised to see his work on the White Mountains National Park, as I’ll be doing a fall roadtrip with my parents there this month! He reimagined a tourist friendly rendition of the White Mountains National Park with an extra emphasis on the National Park and ambulatory landmarks. I’m also inspired by his <a href="https://andywoodruff.com/posts/2023/30-day-map-challenge/"><em>30 Day Map Challenge 2023</em></a>. I might try and run it in Nov 2025.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cS8W_H5pJzbQOmeUDPMstg.png" /><figcaption>Source: <a href="http://White Mountains &quot;National Park&quot; Map">Andy Woodruff’s blog post, <em>White Mountains “National Park” Map</em></a></figcaption></figure><p>Another modern-day cartographer, <strong><em>Daniel P. Huffman</em></strong><em>,</em><strong><em> </em></strong>runs a cheeky website called <a href="https://somethingaboutmaps.com/">somethingaboutmaps.com</a>. As a New Yorker, I love the meticulous, monochromatic map of the Bowery area in Manhattan. Look at the line work below! He’s also made maps across the world, including China, Europe, and India.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XvZkTXRbJAuosKdgmPzkkw.png" /><figcaption><a href="https://somethingaboutmaps.com/The-Bowery">Source</a></figcaption></figure><p><strong><em>James Cheshire </em></strong>is a Professor of Geographic Information &amp; Cartography and Director of the University of College London (UCL) Social Data Institute. Author of eccentric, quirky books, he usually focuses on one topic and makes a plethora of maps on it. For example, <em>London: The Information Capital</em> offers a unique data-driven vantage point of one of the greatest cities in the world.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZLOcp-kTtdLqOFak.jpg" /><figcaption>Source: <a href="https://jcheshire.com/featured-maps/london-information-capital/">James Cheshire’s Website</a></figcaption></figure><p>As an aspiring GIS technologist, this was a great MOOC to learn about cartographic best practices and design thinking that goes into building a map. One key learning was that every decision you take with creating a map is an intentional choice. You’re almost like a director, having to make tedious decisions, so don’t take that responsibility lightly!</p><p>I completed this Esri MOOC just in time, receiving my <a href="https://www.esri.com/en-us/about/ypn/overview">Esri Young Professional Network</a> (YPN) Content Ambassador badge days after. I’d recommend joining this community for those interested in GIS, geospatial modeling, or visual story-telling! Esri YPN also hosts a bunch of events where you can mingle and network with young professionals and industry leaders in the GIS field. Just last week, I attended an Esri YPN event in New York City, and I got to meet cool folks working on GIS projects within the public and private sector.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5B43ic4HBMQDPDRt7q6r4Q.png" /></figure><p>Esri YPN also runs a <a href="https://community.esri.com/t5/esri-young-professionals-network-blog/bg-p/esri-young-professionals-network-blog">blog</a>, a neat forum to learn about the latest in geospatial technology and solve your queries/bugs. I had the opportunity to write a blog post about my experiences navigating the Esri Developer &amp; Technology Summit as a student assistant and attendee. You can check it out <a href="https://community.esri.com/t5/esri-young-professionals-network-blog/navigating-the-esri-developer-amp-technology/bc-p/1656425#M1048">here</a>!</p><p>If you are interested in joining an Esri MOOC, the <a href="https://www.esri.com/training/catalog/5d76dcf7e9ccda09bef61294/spatial-data-science-the-new-frontier-in-analytics/"><em>Spatial Data Science: The New Frontier in Analytics</em></a><em> </em>is running right now. The next one I’ll be enrolling in is <a href="https://www.esri.com/training/catalog/68250ef99d321b7e769b5e9f/the-arcgis-imagery-mooc-foundations-and-frontiers/"><em>The ArcGIS Imagery MOOC: Foundations and Frontiers.</em></a> Maybe, I’ll catch you there. Cheers xx</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f0c4040f1d01" width="1" height="1" alt=""><hr><p><a href="https://medium.com/between-the-graticules/notes-on-esris-cartography-mooc-f0c4040f1d01">My Favorite Cartographers of All Time</a> was originally published in <a href="https://medium.com/between-the-graticules">Between the Graticules</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Notes on 2025 Build with Mapbox]]></title>
            <link>https://medium.com/between-the-graticules/notes-on-2025-build-with-mapbox-0eba2a376a86?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/0eba2a376a86</guid>
            <category><![CDATA[maps]]></category>
            <category><![CDATA[gis]]></category>
            <category><![CDATA[cartography]]></category>
            <category><![CDATA[built-with-mapbox]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Mon, 15 Sep 2025 05:13:56 GMT</pubDate>
            <atom:updated>2025-10-24T14:32:01.237Z</atom:updated>
            <content:encoded><![CDATA[<p>This past week, Mapbox hosted its annual developer summit, <em>Build with Mapbox</em>. Across keynotes, technical sessions and demos, the company unveiled new developer tools, product enhancements, and emerging AI strategies. With the tech industry moving at an almost breathless pace, it feels important not only to take stock of technical progress but also consider its ripple effects in GIS, particularly in navigation, situational awareness, and cartographic storytelling. What follows are my notes and reflections on three themes from the summit, along with a few hypotheses about where <em>GeoAI </em>may be heading.</p><p>The first theme centers on enhancements to Mapbox’s bespoke cartographic tools, making navigation and visualization more immersive and interactive. In response to developer and users feedback on 3D features, Mapbox has rolled out procedural building coverage, a technique that uses machine learning and styling rules to generate assets at run-time in lieu of pre-made models. The result is a more dynamic and interactive cartographic experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/954/1*TrlT_1Zd0zG72dKce5pZfg.png" /><figcaption>Snippet of video demo from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/PWl-X4KfTkqCYAOBkBPZAg">What’s New in Mapbox Navigation: Build 2025 Highlights</a> session, presented by <a href="https://www.linkedin.com/in/mleal/">Marco Leal</a>, Director of Product Navigation at Mapbox</figcaption></figure><p>In a demo of 3D live navigation and lane guidance, <a href="https://www.linkedin.com/in/mleal/">Marco Leal</a>, Director of Product Navigation, shows a car immersed in a scan of the built environment and surrounding vehicles. As I watched the demo, I noted Mapbox’s preference for synthetic, stylized renderings over photorealistic depictions of the built environment. This design choice creates a more gamified experience that effectively augments reality. Put another way, users, immersed in this mixed-reality medium, feel present within the city rather than viewing it from a conventional bird’s-eye vantage. I imagine that, beyond human users, this augmented situational awareness has implications for autonomous vehicle research: as cars’ awareness of their surroundings is heightened, could they make more nuanced decisions on the road?</p><p>Beyond navigation, the storytelling and world-building potential of customizable maps is hard to ignore. I was particularly impressed by <a href="https://www.linkedin.com/company/cecilianpartners/posts/?feedView=all">Cecilian Partners</a>, a proptech company, that utilizes Mapbox to design bespoke digital maps for real estate projects. Their “map stack” layers topography and local amenities into a single immersive experience. The visual below is the epitome of how stylized cartography can merge functionality with narrative.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BUDKrcpKlZpnUTAc3j68HA.png" /><figcaption>Screenshot from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/jF-lQdiJQcCMs7iSPE-9Kw/video/UWfTkrh9QcKQKr6CkutfdQ">The art of place: Maps that merge design, data, and discovery</a> session, presented by <a href="https://www.linkedin.com/in/luke-h-decker/">Luke Decker</a>, Director of Visual Services at <a href="https://www.linkedin.com/company/cecilianpartners/posts/?feedView=all">Cecilian Partners, Inc.</a></figcaption></figure><p>The next theme is the rapid advancement of agentic workflows within GIS. In keeping with the tech industry, Mapbox has launched a <a href="https://github.com/mapbox/mcp-server">Model-Context-Protocol (MCP) server</a> and <a href="https://github.com/mapbox/mcp-devkit-server">MCP DevKit</a>. The MCP server provides a standardized framework for calling Mapbox APIs, transforming any AI agent or application into a geospatially aware system, while the DevKit helps developers better build and customizable maps. The location agent unifies these components, orchestrating multi-agent workflows through text-based prompting.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/930/1*__uQyftBPoL1IAN5KhesZg.png" /><figcaption>Live demo from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/oNHe-dgcRWycEvX36P_qhg">Build an agent with Mapbox MCP</a> session, presented by Valentin Salmanovicius, ML Engineer</figcaption></figure><p>Valentin Salmanovicius demonstrated building a GeoAI agent using Mapbox’s MCP. Within an hour, he created a map integrated with a chatbot! The chatbot was so adept geographically that it was a bit vexing: it could responded to location-based questions, dynamically panned to points of interest, and generated isochrones and travel-time matrices. While latency was noticeable at times, the demo was emblematic of the potential of this infrastructure for real-time map reasoning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vvHYsxyMFkfNvkMN9aTMlg.png" /><figcaption>Slide from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/QJiJZ6A1RA27U6hyexG2HA">Artificial Intelligence, meet location intelligence</a>, presented by <a href="https://www.linkedin.com/in/anu-sharma-039b451/">Anu Sharma</a>, VP of Product at Mapbox</figcaption></figure><p><a href="https://www.linkedin.com/in/anu-sharma-039b451/">Anu Sharma</a>, VP of Product at Mapbox, took Salmanovicius’ demo a step further. In her session, an enhanced chatbot sits alongside a map, helping users explore their surroundings and book hotels/restaurants within a single interface. This integration illustrates the real power of MCP and DevKit! I find that online excitement sometimes treats the tools themselves as an end rather than a means. So to me, Salmanovicius’ MCP demo is most compelling in the context of Sharma’s reflections on the possibilities of combining artificial intelligence with location intelligence.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PO8iRKUa99YXiBqUHV46Cg.png" /><figcaption>Slide from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/QJiJZ6A1RA27U6hyexG2HA">Artificial Intelligence, meet location intelligence</a>, presented by <a href="https://www.linkedin.com/in/anu-sharma-039b451/">Anu Sharma</a>, VP of Product at Mapbox</figcaption></figure><p>The third theme that emerges from agentic GeoAI is less a definitive answer than an open question: what is the most effective input medium for AI assistants that sit alongside maps? Mapbox engineer Antonio Zugaldia unveiled the new AI-powered, voice-driven feedback agent. With this agent, users can report traffic issues, building inaccuracies, and car connectivity glitches entirely via voice.This minimizes distractions, offering a safer experience for the user. Moreover, multiple reports can be submitted in one session, because LLMs and AI can now handle dirty data effectively. All of this offers such rich and nuanced reporting metrics and logs for developers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0ZpqN8ik6KuMg5fh8LvtAA.png" /><figcaption>Snippet of video demo from <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/xCkQzyq9Txa74jaR7krLCw/video/GdRChmAjSna9sRGIv_QLgw">Mapbox Feedback Agent: Enable voice-driven feedback in applications</a> presented by Antonio Zugaldia, Engineering Manager of Navigation at Mapbox</figcaption></figure><p>To me, this approach shows that text-based prompting, common in chatbots, is not always ideal for maps. Maps already rely on high-fidelity visual information. Overloading this visual medium with textual input can be counterproductive; voice offers a more productive and intuitive interaction. Combining the voice-enabled Feedback Agent with the capabilities demonstrated in Sharma’s demonstration unlocks multi-step, voice-driven navigation: imagine asking an assistant to find the closest grocery on the way from work and locate economical petrol stations i.e. conversationally planning a complex route in real time. This kind of interaction suggests that voice may be a more productive medium for interacting with maps, than say, text-based prompts.</p><p>With a busy school schedule, I was glad to catch Mapbox’s developer conference! This week illustrated Mapbox’s commitment to delivering best-in-class cartographic capabilities to clients, developers, and users. In the vignettes of public–private partnerships — with speakers from TripAdvisor to BMW to FlightPulse — I was impressed by how experiments with industry and community partners are, time to time, folded back into the core platform, extending their reach to the wider Mapbox ecosystem.</p><p>I personally have been an ardent fan of Mapbox, having used it in two personal projects, <a href="https://nyceats.live/">NYC Eats</a> and <a href="https://atmikapai.dev/">my cartography portfolio</a>. The theme of storytelling and world-building has always been provocative and compelling to me. Naturally, I drew plenty of inspiration from the <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/fsLV1xwGTeqee920_Uc5Ng"><em>Customize Mapbox Standard in Studio</em></a> and <a href="https://events.zoom.us/ejl/AqJYeHLi-yS_9Q7FPvPQ8gh7Txi_A3tU3R3nplxvTTfz9ntk-S_L~Azn1wugC0u4Jqkwg-gtyZpG-7D-EZDYuv-wV0XcHFWr8zpwGwP-ytrfsDYM3zMUQK1h-o1U0yCNVNEzql-z_-H9lD/sessions/jF-lQdiJQcCMs7iSPE-9Kw"><em>The Art of Place: Maps that Merge Design, Data, and Discovery</em></a> sessions. Some of these ideas have already found their way into my portfolio. Check it out at <a href="https://atmikapai.dev/">atmikapai.dev</a>. Next, I am keen on exploring Mapbox’s MCP server for curated map × chatbot integrations — stay tuned!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0eba2a376a86" width="1" height="1" alt=""><hr><p><a href="https://medium.com/between-the-graticules/notes-on-2025-build-with-mapbox-0eba2a376a86">Notes on 2025 Build with Mapbox</a> was originally published in <a href="https://medium.com/between-the-graticules">Between the Graticules</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[NYC Runs]]></title>
            <link>https://medium.com/map-nyc/nyc-runs-99234a716c36?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/99234a716c36</guid>
            <category><![CDATA[arcgis]]></category>
            <category><![CDATA[maps]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[google-maps]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Wed, 06 Aug 2025 15:11:32 GMT</pubDate>
            <atom:updated>2025-09-15T01:37:59.249Z</atom:updated>
            <content:encoded><![CDATA[<h3>NYC Races</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*umUdi5Zg8OgKRL9Wx6hDDg.png" /><figcaption><a href="https://atmikapai13.github.io/iconic-runclubs-nyc/">Website</a></figcaption></figure><p>During Open Data Week 2025, I was inspired by Dan Dewitz’s <em>Mapping Iconic NYC Sandwiches</em>, a brilliant cartographic homage to a New York Times-curated list. His talk reminded me how geographic data can illuminate not just infrastructure but identity and culture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Fdtp2gx8R3-eNU43WoJfCg.png" /><figcaption><a href="https://www.iconicsandwiches.com/">Iconic NYC Sandwiches Mapped</a></figcaption></figure><p>As a runner, I decided to make a healthier tribute: Iconic NYC Run Clubs.</p><p>The dataset began with a community-curated Google Sheet compiled by NYC native runner <a href="https://www.instagram.com/shaylaruns">@shaylaruns</a>, listing dozens of run clubs across boroughs. I used Google Places API to geotag each location and then visualized the results using ArcGIS Java SDKs, tapping into the spatial DNA of the city.</p><p>This is data that wants to be mapped, networks of movement, gathering, and sweat woven through the city grid. NYC Runs is a kinetic counterpoint to static civic datasets. It maps New York by how it breathes.</p><p>Explore the map and maybe join a run!</p><p><a href="https://atmikapai13.github.io/iconic-runclubs-nyc/">NYC Run Clubs Map - ArcGIS</a></p><p>Also, check out <strong>NYC Eats</strong>, where I map some of the city’s best culinary spots, from classic delis to hidden gems. Together, these projects offer a layered view of New York’s vibrant culture.</p><p><a href="https://nyceats.live/">NYC Eats - Restaurant Week Map</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=99234a716c36" width="1" height="1" alt=""><hr><p><a href="https://medium.com/map-nyc/nyc-runs-99234a716c36">NYC Runs</a> was originally published in <a href="https://medium.com/map-nyc">The New York Trilogy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[NYC Eats]]></title>
            <link>https://medium.com/map-nyc/nyc-eats-724f1ae31a91?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/724f1ae31a91</guid>
            <category><![CDATA[cursor]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[web-scraping]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Mon, 28 Jul 2025 14:57:55 GMT</pubDate>
            <atom:updated>2025-08-06T15:13:28.168Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JiMx_Jbocsbats1CP5tKYA.png" /></figure><p>It’s NYC Restaurant Week! I built a <a href="https://nyceats.live/">web app</a> to help NYC foodies find the best deals. The interactive map visualizes all participating restaurants and overlays curated lists like the Michelin Guide, Bib Gourmand, and the NYT Top 100, making it easier to spot standout Restaurant Week offerings.</p><p>This data is inherently geographic, but the current NYC Tourism site doesn’t fully reflect that. The current implementation displays over 600 restaurants across more than 50 pages, with 12 listings per page, making it difficult for users to browse efficiently and identify the best deals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gvdcL2-VpSPHbmQ-N8XL_g.png" /><figcaption>Parsing through 53 pages of restaurants is not the most user-friendly. Source: <a href="https://www.nyctourism.com/restaurant-week/">NYC Tourism</a></figcaption></figure><p>Building a map felt like the natural next step and a useful one for foodies like myself. I also wanted to bring together menus, hours, and booking information in one place with a more user-friendly UI/UX. I hope it helps you make the most of Restaurant Week. Feel free to share with friends! 🥢 (Read more <a href="https://www.linkedin.com/feed/update/urn:li:activity:7353847173093994497/">here</a>!)</p><p><strong>Tech stack:</strong> React, TypeScript, Python for web scraping, Mapbox GL JS</p><p><a href="https://nyceats.live/">https://nyceats.live/</a></p><p><a href="https://nyceats.live/">NYC Eats - Restaurant Week Map</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=724f1ae31a91" width="1" height="1" alt=""><hr><p><a href="https://medium.com/map-nyc/nyc-eats-724f1ae31a91">NYC Eats</a> was originally published in <a href="https://medium.com/map-nyc">The New York Trilogy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Notes on 2025 Esri Developer & Technology Summit]]></title>
            <link>https://medium.com/between-the-graticules/notes-on-2025-esri-developer-technology-summit-75cd4b496d5d?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/75cd4b496d5d</guid>
            <category><![CDATA[gis]]></category>
            <category><![CDATA[urban-planning]]></category>
            <category><![CDATA[esri]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Sun, 06 Apr 2025 19:04:49 GMT</pubDate>
            <atom:updated>2025-10-24T14:28:18.942Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ji09q9Rg_tf60CLRIoCc7A.png" /></figure><p>The Environmental Sciences Research Institute (Esri) hosts four major GIS summits each year as the global leader in GIS technology. I had the opportunity to attend the 2025 Esri Developer &amp; Technology Summit in Palm Springs, California, as a student assistant. As an aspiring GIS developer, this was an awesome opportunity to learn more about emerging GIS technologies and the industry’s strategies on nascent technologies like AI and LLMs.</p><p>The four-day summit began with plenary sessions introducing Esri’s latest product releases, interspersed with live demos showcasing cutting-edge advancements. This was followed by technical sessions covering key GIS topics and an expo for Esri’s extensive suite of tools, including ArcGIS Pro, Game Engine integrations, ArcGIS Earth, and more.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*occd4NzdKT5ocS1ca2ZQIw.png" /></figure><p>One of the first plenary sessions I attended was one led by Michael Davidson and Jim McKinney, which focused on 3D GIS. They demonstrated how ArcGIS Geo and Digital Twins are transforming spatial analysis, enabling stakeholders to create and interact with digital representations of real-world environments. I had been introduced to digital twins at the 2024 Urban Tech Summit at Cornell Tech, but McKinney’s session showed me how governments and public sector stakeholders are thinking about digital twins.</p><p>That said, I remain somewhat skeptical of their practical application— I still struggle to distinguish the hype from real-world use cases. I am also more drawn to augmented and mixed reality (AR/VR) spaces than a virtual replica of a physical environment. Of course, AR/VR relies on having an accurate representation of the built environment — like a digital twin — to function effectively.</p><p>(Here is a cool <a href="https://storymaps.arcgis.com/stories/b43b415b07ff4f858f8091e9bbca9810?adusf=linkedin&amp;aduc=esri&amp;adut=e30d1d94-19a3-4fa2-a725-9bfc014526f7">showcase</a> of how the digital twin of Zürich in Switzerland was created by Esri developers and how it can be leveraged through various applications.)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/958/1*yU6tJeBxki-7LDKhpwmwYA.png" /></figure><p>Another popular session had to do with Esri’s AI strategies, delivered by Sud Menon, the Director of Products &amp; Technology at Esri.<br>He broached the idea of AI assistants and natural language processing (NLP) integrations into GIS workflows. To that effect, his colleagues, Rohit Singh and Ishmael Chivite, delivered a demo — Chivite typed simple commands that large language models (LLMs) translated into Arcade code for more advanced dataset filtering. The presenters believed that incorporating LLMs into the ArcGIS Pro workflow would simplify GIS tasks for developers and make spatial data more accessible to users without a strong technical background.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qfq3n-5RLz3QgVv_FaxQ3Q.png" /><figcaption>Rohit Singh delivering a live demo on the use of LLMs within Esri’s ArcGIS Pro</figcaption></figure><p>While, in theory, this is an altruistic and noble goal, I remain unconvinced of its practical value. Developers already use tools like ChatGPT and Claude for rudimentary coding tasks, and non-technical stakeholders — such as product managers — can often delegate these tasks. So, what is the real purpose of embedding NLP into GIS workflows?</p><p>One could argue it is beneficial to neophyte GIS users, who can learn within the ArcGIS Pro platform with the help of NLP/LLMs without having to exact out of the application. This would reduce the learning curve, which is a sticking point for Esri’s technologies. Even in this scenario, I would push back, questioning how much value is NLP integration actually bringing to ArcGIS Pro workflows. This leads me to a broader question: does every company really need to develop its own LLM or chatbot — and is there genuine value in doing so?</p><p>On the one hand, companies are either following the bandwagon or intentionally developing these tools in anticipation of the future. (Whether the motivation is strategic or just sheep mentality is a moot point — better to be ready than risk falling behind.) If Parag Agrawal’s efforts at Parallel Web Systems come to fruition, agentic assistants from various companies will be able to communicate and collaborate in a “fifth” dimension. In that case, building functional LLMs is the necessary first step to process the kind of data these assistants will be predisposed to, and kudos to Esri for taking that step by integrating LLMs into their workflows.</p><p>On the flip side, if Esri and the like are simply following the hype, then there may not be much tangible in-house innovation happening. It starts to feel like non-tech companies are just repackaging LLMs developed by tech giants rather than pushing the frontier. This makes me wonder: how can LLMs truly advance GIS beyond helping with rudimentary code conversions or basic mapping tasks? What if I could have a meaningful conversation with my GIS assistant — not just about syntax, but about planning, spatial ethics, or scenario modeling? I would urge Esri to explore such projects, given their expertise in the field, their strong partnerships with both the public and private sectors, and the entrepreneurial foresight of Jack Dangermond, the founder and CEO of Esri, in putting GIS on the commercial map (pun intended).</p><p>All of that is to say, the sessions were incredibly thought-provoking, leaving ample room for further exploration. I left the event with at least ten ideas on urban planning and geospatial analysis, with the intention of leveraging Esri’s rich ecosystem of GIS tools. Beyond the technical sessions, another valuable aspects of the summit was the diverse mix of attendees. I connected with Esri developers from Italy and Australia, and I was truly in awe of the company’s global footprint. Many of these developers have been attending the summit for years and eagerly anticipate it each year!</p><h3><strong>What is an Esri Student Assistantship?</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uQGEBe-MYxLNXlFBFZ56gg.jpeg" /><figcaption>2025 Esri Developer and Technology Summit Student Cohort</figcaption></figure><p>Speaking to my experiences as a student assistant, I got to meet 18 other undergraduate and graduate students from across U.S. Together, we covered myriad fields of study from information science to geography and even archeology. Our responsibilities as student assistants encompassed setup, attendee registration, managing the Esri store, assisting with technical sessions, providing audio/video support, and more. Even then, there was plenty of time to network with industry professionals and and attend presentations, workshops, and the Expo.</p><p>Overall, my experience as a student assistant at the ESRI Developer &amp; Technology Summit was incredibly rewarding. If you’re passionate about GIS, spatial analytics, or the intersection of technology and public-private sector partnerships, I highly recommend volunteering for Esri’s largest conference, the User Conference taking place July 2025 in San Diego — you might catch me there!</p><p>Check out my post on <a href="https://community.esri.com/t5/esri-young-professionals-network-blog/navigating-the-esri-developer-amp-technology/ba-p/1653723?aduc=Social">Esri’s YPN blog</a>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=75cd4b496d5d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/between-the-graticules/notes-on-2025-esri-developer-technology-summit-75cd4b496d5d">Notes on 2025 Esri Developer &amp; Technology Summit</a> was originally published in <a href="https://medium.com/between-the-graticules">Between the Graticules</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Disaster Detection Model Using NLP]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/between-the-graticules/building-a-disaster-detection-model-using-nlp-009a11cfc0ef?source=rss-63b28313683a------2"><img src="https://cdn-images-1.medium.com/max/648/1*eSRehgX3QWF_dmls9cmdPw.png" width="648"></a></p><p class="medium-feed-snippet">In a world connected by tweets and real-time updates, the need to efficiently identify genuine disaster-related messages has never been&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/between-the-graticules/building-a-disaster-detection-model-using-nlp-009a11cfc0ef?source=rss-63b28313683a------2">Continue reading on Between the Graticules »</a></p></div>]]></description>
            <link>https://medium.com/between-the-graticules/building-a-disaster-detection-model-using-nlp-009a11cfc0ef?source=rss-63b28313683a------2</link>
            <guid isPermaLink="false">https://medium.com/p/009a11cfc0ef</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[urban-planning]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[urban-technology]]></category>
            <dc:creator><![CDATA[Atmika Pai]]></dc:creator>
            <pubDate>Sun, 24 Nov 2024 16:49:30 GMT</pubDate>
            <atom:updated>2025-10-24T14:36:34.888Z</atom:updated>
        </item>
    </channel>
</rss>