<?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[Sequence Blog - Medium]]></title>
        <description><![CDATA[Sequence is a straight-forward outsourcing service for all your data labelling needs. Gain more accuracy, completeness, and uniformity across all your datasets. - Medium]]></description>
        <link>https://medium.com/sequencework?source=rss----ef452b256034---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Sequence Blog - Medium</title>
            <link>https://medium.com/sequencework?source=rss----ef452b256034---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 03:03:29 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/sequencework" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com" rel="hub"/>
        <item>
            <title><![CDATA[Build a New Generation of Stores]]></title>
            <link>https://medium.com/sequencework/build-a-new-generation-of-stores-5ca05789e8a1?source=rss----ef452b256034---4</link>
            <guid isPermaLink="false">https://medium.com/p/5ca05789e8a1</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-outsourcing]]></category>
            <category><![CDATA[dataannotations]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[customer-success-stories]]></category>
            <dc:creator><![CDATA[Sequence]]></dc:creator>
            <pubDate>Thu, 08 Nov 2018 07:59:12 GMT</pubDate>
            <atom:updated>2018-11-08T14:17:45.345Z</atom:updated>
            <content:encoded><![CDATA[<p>Learn how <a href="http://storelift.co">Storelift</a> leverages image annotation for computer vision in creating new commercial retail experiences.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dnLODoVy9rHkagGYPY1QzA.png" /></figure><blockquote>“By outsourcing the data annotation to <a href="https://sequence.work/">Sequence</a>, we experienced huge gains in terms of time and quality of work. Indeed, we have been able to reallocate 70% of our time to higher-value tasks.</blockquote><blockquote>Sequence’s expertise enabled us to train our neural networks on clean data, which is crucial for getting the highest accuracy possible. We finalised the MVP of our autonomous cashier system and are now in talks with major European retailers for pilots in 2019”.</blockquote><blockquote>— David Gabai, COO &amp; Co-Founder, <a href="https://www.storelift.co/">Storelift</a></blockquote><p><a href="https://www.storelift.co/"><strong>Storelift</strong></a> is a deep tech startup working on computer vision applications for retail. They are developing an autonomous cashier system that allows retailers to offer a checkout-free experience to their customers in store.</p><p>Storelift’s computer vision technology is designed to automatically track in-store clients and recognise products during the shopper journey to attribute the right product to the right customer.</p><p>To achieve high levels of precision and accurate results, the Storelift data science team has to feed their neural networks with an immense amount of training data.</p><h4>Problem Statement</h4><p>Storelift had generated hours of shopping scenes in-store and split them into individual frames. They then laboriously annotated each person and each product in each frame. By doing so, they were able to map the coordinates of a person or product on an image. It would then indicate to the algorithm that an area on the image is of a person or a product.</p><p><strong>This is an important but extremely time-consuming part of their data science team’s responsibilities.</strong> The process can take up to 70% of their time and often takes away the team’s attention from other important tasks. Annotating the training data in-house was cost-ineffective and tiresome for the team.</p><h4><strong>Solution</strong></h4><p>Together, we discussed which solution would best fit their needs. Storelift needed accurate annotations to get optimal performance and a quick turnaround time. They insisted in particular on the annotation precision. Even a few small errors can generate too much noise and dramatically reduce the algorithm prediction score.</p><p>At Sequence, our in-house technology allows us to work quickly and adapt to the needs of our clients. With a few adjustments, we were able to provide a project-specific annotation interface that fit their needs. This allowed our team of contributors to have more detail when annotating the images and gain more overall accuracy.</p><p><strong>Within a couple of weeks, we were able to annotate over 50,000 images with roughly 100,000 annotations across all images whilst providing high-quality results and excellent accuracy.</strong></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Ffandai.donguapan.workers.dev%3A443%2Fhttps%2Fwww.youtube.com%2Fembed%2FtqPYwF_wLR8%3Ffeature%3Doembed&amp;url=https%3A%2F%2Ffandai.donguapan.workers.dev%3A443%2Fhttp%2Fwww.youtube.com%2Fwatch%3Fv%3DtqPYwF_wLR8&amp;image=https%3A%2F%2Ffandai.donguapan.workers.dev%3A443%2Fhttps%2Fi.ytimg.com%2Fvi%2FtqPYwF_wLR8%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="640" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b2b6f768ab76898e2f1338cab3ed85cd/href">https://medium.com/media/b2b6f768ab76898e2f1338cab3ed85cd/href</a></iframe><h4><em>Learn more about us at </em><a href="https://sequence.work/"><strong>sequence.work</strong></a><strong><em>. Sequence provides annotation and classification outsourcing for data science teams. </em></strong>Gain more accuracy, completeness, and uniformity across all your datasets.</h4><h4>Curious about our image annotation interface? See how it runs here: <a href="https://sequence.work/demo/annotationv2">Annotation Tool Demo</a>.</h4><p><em>Looking to work on tasks instead? </em><a href="https://sequence.work/worker/signup"><strong><em>Become a contributor here</em></strong></a><strong><em>.</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5ca05789e8a1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sequencework/build-a-new-generation-of-stores-5ca05789e8a1">Build a New Generation of Stores</a> was originally published in <a href="https://medium.com/sequencework">Sequence Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Annotation Tool Part 2: Retrospective on the Build]]></title>
            <link>https://medium.com/sequencework/annotation-tool-part-2-retrospective-on-the-build-50b90ff90388?source=rss----ef452b256034---4</link>
            <guid isPermaLink="false">https://medium.com/p/50b90ff90388</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[custom-annotation]]></category>
            <category><![CDATA[product-design]]></category>
            <category><![CDATA[outsourcing]]></category>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Mabel Tan]]></dc:creator>
            <pubDate>Wed, 03 Oct 2018 02:04:43 GMT</pubDate>
            <atom:updated>2020-02-04T13:34:19.463Z</atom:updated>
            <content:encoded><![CDATA[<p>This article is Part 2 of our series on the <a href="https://sequence.work/demo/annotationv2"><strong>Sequence Annotation Tool</strong></a>. If you’d like to see our objectives for building our own Annotation tool, read it here: <a href="https://medium.com/sequencework/part-1-why-we-built-an-annotation-tool-d1711a51f40b"><strong><em>Annotation Tool Part 1: Why We Built Our Own Tool</em></strong></a><strong><em>.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3ohtn6hwFSxsp6AEGuEIgg.png" /><figcaption>Bounding box annotation using Sequence Annotation Tool</figcaption></figure><p>This article is a look back on our Annotation Tool build. We learnt a lot from it and thought we’d share our process. Here we’ll take a look at what we built and the insights that drove our decision making.</p><p>Let’s start at the beginning.</p><h3>Version 1 — The Prototype</h3><blockquote>“The best tool is the one that does the job.”</blockquote><p>We released Version 1 with minimal but essential features. Our intention at this stage was to serve our earliest customers. We were still exploring the viability of our annotation line of services and wanted to go slow and steady.</p><h4><strong>We started with:</strong></h4><ol><li><strong>Create new polygon/bounding boxes point by point <br></strong>An obvious necessity. No polygons + no bounding box = no annotations.</li><li><strong>Move existing points along the path<br></strong>We knew we’d require some flexibility and leeway for errors. Being able to adjust and tweak the shape of the path would be basic.</li><li><strong>Delete existing shapes</strong> <br>Errors are inevitable. We wanted a tool that was forgiving enough for the user even though it was just an MVP.</li></ol><h4><strong>We did not include:</strong></h4><p><strong>Zoom and pan<br></strong>We knew intuitively that this would be something our contributors would need. However, it would have extended the build and delayed our time to start the project. We ran with the browser zoom and kept our fingers crossed. 🤞</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r1c999-dETgErsVYJ7m7jg.png" /><figcaption>The primitive version of our Annotation Tool 😅</figcaption></figure><p>Thankfully, the outcome was that the annotation tool worked for what we required at the time <em>*collective sigh of relief*</em>.</p><p><strong>However, a few frustrating factors quickly became apparent.</strong> We ran some usability testing with our contributors, and we gathered an enormous amount of feedback for Version 1. The pain points were clear.</p><h4><strong>The feedback collected highlighted a few issues:</strong></h4><ol><li>Initial points on the path were too large when zoomed-in making fine-tuning nigh impossible.</li><li>Inability to click through a point to the point beneath it to close the polygon path.</li><li>Zooming in and out using the browser did not produce an image that scaled proportionally to the window.</li><li>Lacking tools to reduce errors — <em>‘undo’</em> and <em>‘delete’ </em>points were unavailable making it difficult and tedious for contributors to correct mistakes — often resulting in the need to entirely delete and redraw shapes.</li><li>A few of our contributors needed tablet support <em>(annotating images using your finger on your tablet is unsurprisingly a troublesome adventure)</em>. However, this group is a minority use case, so we needed to aid them but in a minimal but effective way.</li><li>The view was also cluttered and not responsive to browser dimensions.</li></ol><h4>All these insights allowed us to pinpoint our issues and understand our contributors better. We took all this feedback and injected it into our solutions for Annotation Tool Version 2.</h4><h3>Version 2 — Improving Precision, Error Reduction, and Usability.</h3><p>Our goal for Version 2 was to reduce the likeliness of error occurrence and significantly improve the usability of the tool. Our insight being that even though we can’t control a person’s capacity to maintain accuracy, we could control the tools we provided that would help reduce or correct mistakes, and by doing so we could improve overall precision.</p><p>We approached Version 2 — armed with our insights. <strong>We wanted to create a power tool that was applicable for a general audience;</strong> so we tried to bring over some of the interactions and behaviours that we found useful in professional software. Therefore, a lot of the ideas around the new functionalities are referenced from other vector software like Photoshop and Sketch.</p><h4>Version 2 feature implementations:</h4><p><strong>1. Zoom and pan<br></strong>This was our most requested feature. Previously when zoomed in with the browser, the image would not scale linearly and you were not able to add more detail. Zoom and pan come as a pair because while working zoomed-in refines output, when zoomed-in too far, you lose a sense of the bigger picture. Being able to pan allows you to refine and maintain context of the surrounding image — this allows you to work more comfortably for extended durations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*Yfe5vj4xM7fWYUxrXDGLOA.gif" /><figcaption>Zoom and pan for image annotation tool</figcaption></figure><p><strong>2. Visual cues <br></strong>Cursor events that indicate different functionalities on hovering over different areas.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/784/1*BB6oTmJo7DT8AgkW_ZGctQ.png" /><figcaption>Cursors cues for: closing a path, adding a point, and panning the image respectively.</figcaption></figure><p><strong>3. Add/Delete/Move point on a path<br></strong>Being able to add or remove point detail on a path to get that curved edge <em>just right</em> was important for us. In addition, being able to undo a previous point without having to redraw the polygon was a basic necessity.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/312/1*bgqFAFj5ixs8cqUxu-zxgQ.gif" /><figcaption>Adding a point to a polygon annotation</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/312/1*xXz89qPYs-T7ZtqMc4uTJw.gif" /><figcaption>Deleting a point on a polygon annotation</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/312/1*aM2qKN7ITWE5qcTfhog-Lw.gif" /><figcaption>Moving and repositioning a point on a path for a polygon annotation</figcaption></figure><p><strong>4. Cursor snapping<br></strong>There was a frequent point of frustration with overlapping points whereby it was difficult to find the starting point for the polygon and close the polygon shape. Being able to snap to the first point, reduces wasted actions and minimises unnecessary frustration.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/312/1*68rO7ruTBHFfOrbDfybJAA.gif" /><figcaption>Cursor snapping to close a polygon annotation</figcaption></figure><p><strong>5. Hide annotation layer<br></strong>This was necessary for contributors working on displays with lower resolutions. It reduces overall clutter in the view when working on images with multiple annotations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*MCXg6grFMAppkuqL9gdcuQ.gif" /><figcaption>Hide annotation layer</figcaption></figure><p><strong>6. Annotation shape panel and highlights <br></strong>The shape panel is a functionality that provides organisation and context of work done. It’s an overview of individual shapes that make up all the annotations within the image. In the shapes panel, you can choose to highlight a layer for review before you submit or delete whole layers that you might be unhappy with.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*WTvJwVxg1-6U-ZzkxnCIoQ.gif" /><figcaption>Highlight and delete annotation layer</figcaption></figure><p><strong>7. Accessibility fundamentals and MVP gesture support.<br></strong>Initially, we toyed with the idea of just having keyboard shortcuts instead of buttons. However, we concluded that visual buttons allow contributors to know these actions are possible and enabled functionality, especially on devices that don’t support keyboards.</p><p>We included gesture support ie. pinch to zoom and two-finger touch for panning on touch devices. These were fundamental to ensure that our improved zoom and pan functionality was available to all our users.</p><h4><strong>Overall, we consider Annotation Tool Version 2 a successful build.</strong> We made incremental improvements based on our learnings in Version 1 and they have shown to be impactful for our users in creating a more refined output.</h4><p>We’ve already got some feedback from our contributors and we’re happy to report that the improvements have been well received :)</p><blockquote><em>“I’m working on a project right now. I love the new tools :) Thank you!” — </em><strong><em>Grace</em></strong></blockquote><blockquote><em>“The improvement really helps! The new feature makes it easier to annotate tight spaces as it allows overlapping points.” — </em><strong><em>Amie</em></strong></blockquote><h4>Want to see the Annotation Tool for yourself? Take it for a test drive here: <a href="https://sequence.work/demo/annotationv2">Annotation Tool Demo</a>.</h4><p><em>This article was written in collaboration with </em><a href="https://twitter.com/andyngo"><strong><em>Andy Ngo</em></strong></a><strong><em>.</em></strong></p><p><em>Learn more about us at </em><a href="https://sequence.work/"><strong><em>sequence.work</em></strong></a><strong><em>. We provide annotation and classification outsourcing for data science teams.</em></strong></p><p><em>Looking to work on tasks instead? </em><a href="https://sequence.work/worker/signup"><strong><em>Become a contributor here</em></strong></a><strong><em>.</em></strong></p><p>*Images used in the sample annotations are from <a href="https://unsplash.com/">Unsplash</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=50b90ff90388" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sequencework/annotation-tool-part-2-retrospective-on-the-build-50b90ff90388">Annotation Tool Part 2: Retrospective on the Build</a> was originally published in <a href="https://medium.com/sequencework">Sequence Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Annotation Tool Part 1: Why We Built Our Own Tool]]></title>
            <link>https://medium.com/sequencework/part-1-why-we-built-an-annotation-tool-d1711a51f40b?source=rss----ef452b256034---4</link>
            <guid isPermaLink="false">https://medium.com/p/d1711a51f40b</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[custom-annotation]]></category>
            <category><![CDATA[outsourcing]]></category>
            <dc:creator><![CDATA[Mabel Tan]]></dc:creator>
            <pubDate>Tue, 25 Sep 2018 15:27:54 GMT</pubDate>
            <atom:updated>2018-10-03T06:32:35.611Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This is the first article in a series for our Sequence Annotation Tool. Read on to learn more about our journey along the learning curve.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p3n_qAcsXxm191z8zqVvNw.png" /><figcaption><a href="https://sequence.work/demo/annotationv2">Polygon Annotation of a Cat. Created in Sequence Annotation Tool.</a></figcaption></figure><p>At Sequence, we provide text and image annotation services for machine learning projects. Our most common use cases fall under <strong>image annotation at high volumes</strong>. There are many quality open source annotation tools across the internet and this is often a good solution for most businesses. However, we’ve decided to build our annotation tool in-house. In this article, we’ll talk about some reasons why.</p><p><strong>Annotation is a core functionality of our platform</strong>. Our decision to build our own tool focuses on laying the right foundation for future builds. For us, it’s a win-win situation. By investing into our own annotation tool, we can better serve the needs of our clients and our contributors while working with a tool that’s just the right fit.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*a-fSG7V2hM-AYbYWP5hjoA.png" /><figcaption><a href="https://sequence.work/demo/annotationv2">Bounding Box Annotation of a Dog. Created in Sequence Annotation Tool</a></figcaption></figure><p>We wanted our Annotation Tool to be:</p><h3><strong>1. Web-based</strong></h3><p>We needed the annotation tool to be able to run on the browser. Our platform is web-based, which allows anyone to work on it as long as they have internet access. With no software to install and no manual transfer of work-files from desktop to web, we can simplify the user experience of working on a task.</p><h3><strong>2. Flexible</strong></h3><p>We wanted to be able to configure shortcuts, add specific shapes in the future (beyond bounding boxes and polygons), and tailor the behaviour of the tool based on our business use cases. As most of the time spent on the platform by contributors is on this tool, being able to control the functionality at a granular level gives us an opportunity to refine details and experiment with new ideas.</p><h3><strong>3. Robust</strong></h3><p>To guarantee the same platform behaviour across browsers, and under slower network conditions — we think an in-house build is a right choice for us. As small bugs can sometimes snowball into significant negative experiences within the platform, it was important for us to be able to run extensive tests against the tool and make sure it’s easy for us to ensure the quality we wanted to provide.</p><h3><strong>4. Fast</strong></h3><p>Our goal is to limit delays and lags as much as possible. We pride ourselves on keeping the platform feeling responsive, especially on older machines. Building our annotation tool in-house allows us to make sure that the tool runs smoothly for contributors who have low-CPU computers.</p><p>These constraints helped us to focus on what we wanted for our platform and gain clarity on the requirements we needed to best serve our users.</p><p>We’ve launched the tool on our platform as of September 2018 and the results have been positive. We’ve learned a lot from the process and if you’d like to know more about our process on the build, read the follow-up article here: <a href="https://medium.com/sequencework/annotation-tool-part-2-retrospective-on-the-build-50b90ff90388"><strong><em>Annotation Tool Part 2: Retrospective on the Build</em></strong><em>.</em></a></p><h4><strong><em>Want to see the Annotation Tool for yourself? Take it for a test drive here: </em></strong><a href="https://sequence.work/demo/annotationv2"><strong><em>Annotation Tool Demo</em></strong></a><strong><em>.</em></strong></h4><p><em>This article was written in collaboration with </em><a href="https://twitter.com/lucleray"><strong><em>Luc Leray</em></strong></a><strong><em>.</em></strong></p><p><em>Learn more about us at </em><a href="https://sequence.work/"><strong><em>sequence.work</em></strong></a><strong><em>. We provide annotation and classification outsourcing for data science teams.</em></strong></p><p><em>Looking to work on tasks instead? </em><a href="https://sequence.work/worker/signup"><strong><em>Become a contributor here</em></strong></a><strong><em>.</em></strong></p><p>*Images used in the sample annotations are from <a href="https://unsplash.com/">Unsplash</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d1711a51f40b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sequencework/part-1-why-we-built-an-annotation-tool-d1711a51f40b">Annotation Tool Part 1: Why We Built Our Own Tool</a> was originally published in <a href="https://medium.com/sequencework">Sequence Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Use Data Classification to Find the Perfect Home for Your Customer]]></title>
            <link>https://medium.com/sequencework/using-data-classification-to-find-the-perfect-home-for-your-customer-1d2722aeb9f8?source=rss----ef452b256034---4</link>
            <guid isPermaLink="false">https://medium.com/p/1d2722aeb9f8</guid>
            <category><![CDATA[dataannotations]]></category>
            <category><![CDATA[customer-success-stories]]></category>
            <category><![CDATA[data-preparation]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-classification]]></category>
            <dc:creator><![CDATA[Sequence]]></dc:creator>
            <pubDate>Wed, 29 Aug 2018 10:44:05 GMT</pubDate>
            <atom:updated>2018-11-16T08:20:54.309Z</atom:updated>
            <content:encoded><![CDATA[<p>Learn how Right Home uses Sequence to optimise data classification for their matching algorithm.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_AH7NY3xnvZ2jGqS9h6mag.png" /></figure><blockquote>“<a href="http://sequence.work">Sequence</a> helped us where no other solutions were available. They handled a completely unstructured dataset, and managed to convert the data into tasks and classify the data as per our requirements. They are a resourceful partner for any machine learning project.” <br>- Armelle M., CEO, Right Home.</blockquote><p><a href="http://righthome.fr">Right Home</a> is a real estate agency based in Paris that specialises in house hunting. It’s the perfect solution for those looking to buy an apartment or house in the city. Their customers often already have the requirements in mind but don’t have the time or market knowledge to proceed by themselves. Right Home provides a professional and personal service from the first to the last step of the buying process.</p><h4>Problem Statement</h4><p>Manually searching for homes in a market made of millions of properties is no mean feat. On top of which many of the website are not optimised and have poor search functionality. This all adds up to inefficient work that is extremely time-intensive.</p><p>To overcome this, Right Home uses machine learning to match properties to their client’s requirements. The challenge that then arose was to prepare a large enough data-set for the model. The goal was to categorise the information from tens of thousands of real estate ads into different classes.</p><p><strong>Right Home contacted us to figure out how we could help them to prepare their unstructured data-set.</strong> As all machine learning models do, they needed a high accuracy outcome to ensure the best quality.</p><h4>Solution</h4><p>To provide this quality, we began by on-boarding contributors that had an <strong>accuracy of at least 95%</strong>. Also, as Right Home is a French company, we selected French speaking contributors for the task. We then ran assessments to give them a clear understanding of the real estate industry’s terms and vocabulary before contributors were able to work on tasks.</p><p>To process the unstructured dataset and redistribute the data points as tasks, we build templates that are custom to each project. Our team was able to deliver an intuitive and user friendly interface for our contributors. Allowing them to work with more clarity and less of the clutter from all the unoptimised real estate websites. These templates are a simple way that allows us to save time and money for our clients.</p><p>Close collaboration with Right Home was key to the success of this project. Our Account managers are the point of contact that helped us to guarantee a smooth experience for both the contributors and the client. Navigating questions and hurdles together allowed us to improve the project outcome as it went along. Through this partnership, we categorised 20,000 ads, a total of 200,000 data points.</p><p><strong>Sequence is an outsourcing service for data science teams. Our products include data annotation, tagging, and classification for all project sizes. We offer a free trial to new customers — learn more at our website: </strong><a href="http://sequence.work"><strong>sequence.work</strong></a><strong>.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1d2722aeb9f8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sequencework/using-data-classification-to-find-the-perfect-home-for-your-customer-1d2722aeb9f8">Use Data Classification to Find the Perfect Home for Your Customer</a> was originally published in <a href="https://medium.com/sequencework">Sequence Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Simple Outsourcing for Data Science Teams]]></title>
            <link>https://medium.com/sequencework/crowdsource-human-intelligence-8cbc182c7abe?source=rss----ef452b256034---4</link>
            <guid isPermaLink="false">https://medium.com/p/8cbc182c7abe</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[mechanical-turk]]></category>
            <category><![CDATA[data-preparation]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[computer-vision]]></category>
            <dc:creator><![CDATA[Mabel Tan]]></dc:creator>
            <pubDate>Fri, 20 Apr 2018 08:37:58 GMT</pubDate>
            <atom:updated>2018-08-29T10:53:41.065Z</atom:updated>
            <content:encoded><![CDATA[<p><em>We specialise in data annotation, tagging and classification.</em></p><p>From image annotation to URL labelling, whatever the field that you’re in — be it data science or e-commerce operations — we all need a little help when it comes to processing data at high volumes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wzYm115yB3xqQpl1FZBvfQ.png" /><figcaption><a href="https://sequence.work/use-cases/">sequence.work/use-cases</a></figcaption></figure><p>We’re a small but international team based in France. Our mission is to bring you simple tasks that need human judgement like data labelling, tagging and categorisation at an affordable price with great quality — and all the perks of a managed service.</p><h4><strong>Check out our use cases here: </strong><a href="https://sequence.work/use-cases/"><strong>sequence.work/use-cases</strong></a></h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cbc182c7abe" width="1" height="1" alt=""><hr><p><a href="https://medium.com/sequencework/crowdsource-human-intelligence-8cbc182c7abe">Simple Outsourcing for Data Science Teams</a> was originally published in <a href="https://medium.com/sequencework">Sequence Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>