<?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 Lindsey Whitley on Medium]]></title>
        <description><![CDATA[Stories by Lindsey Whitley on Medium]]></description>
        <link>https://medium.com/@lindsey.whitley?source=rss-7b0cecf1231------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*GxH1bI6U0WLI9Gn4NY_JwQ.jpeg</url>
            <title>Stories by Lindsey Whitley on Medium</title>
            <link>https://medium.com/@lindsey.whitley?source=rss-7b0cecf1231------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 15:25:45 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@lindsey.whitley/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[Service Level Objectives for Engineering On-call Peace of Mind]]></title>
            <link>https://engineering.gusto.com/service-level-objectives-for-engineering-on-call-peace-of-mind-61bb5994d202?source=rss-7b0cecf1231------2</link>
            <guid isPermaLink="false">https://medium.com/p/61bb5994d202</guid>
            <category><![CDATA[continuous-integration]]></category>
            <category><![CDATA[incident-response]]></category>
            <category><![CDATA[best-practices]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Lindsey Whitley]]></dc:creator>
            <pubDate>Thu, 28 Jan 2021 18:40:32 GMT</pubDate>
            <atom:updated>2024-06-25T19:15:45.206Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4MEvG64u6n32uTFjmqeUIg.jpeg" /></figure><p>My team, Developer Experience (DEx), owns the <a href="https://www.atlassian.com/continuous-delivery/continuous-integration">Continuous Integration (CI)</a> experience at Gusto. When I joined the team about a year ago, we were getting pressure from our customers (Gusto’s Product Engineers) to make our CI tool “better”.</p><p>We framed our work to make CI “better” as an experiment with defined questions and metrics from the beginning. Could we use metrics to improve the experienced quality of CI? Once we knew that we were meeting some threshold of quality, could we use these metrics to prioritize future work? We gave ourselves a month to answer the first question and hoped that we’d have a while to test the second.</p><h3>Advantages of where we ended up</h3><p>By defining SLOs and monitors around the performance that we expected from our CI setup in <a href="https://buildkite.com/docs/tutorials/getting-started">Buildkite</a>, we:</p><ul><li>Ensured that we would know about potential degradation in performance before our customers saw it</li><li>Removed the need for a human to look at a dashboard</li><li>Automated prioritization of new features vs “keeping the lights on”</li><li>Gave ourselves a needle that we could see moving up and down with changes, as opposed to needing customer feedback more often</li></ul><h3>Defining “better”</h3><p>We took on the task to understand what issues in CI were actually frustrating Gusto’s engineers. To gather this information, we opened up a couple of channels for feedback.</p><p>We asked for opinions in our #engineering Slack channel. We also scheduled live feedback sessions with product engineers from each group or mission. In these conversations, we asked:</p><ul><li>What is good about our CI?</li><li>Is speed an issue?</li><li>When are you frustrated with our CI?</li></ul><p>Answers to our questions were grouped into buckets, and the most prominent were:</p><ul><li>Builds fail because something goes wrong with docker or mysql or something else outside my control as a Product Engineer.</li><li>I hate when I have to wait for 15 minutes before the build even starts.</li><li>We have so many flaky tests that I assume any <a href="https://rspec.info/">rspec (ruby test suite)</a> failure is caused by them.</li><li>The test suite just takes too long to run.</li></ul><p>With all of this in mind, we knew what “better” would look like for our customers, but we weren’t sure that we were the only team who should be working on solutions to all of those problems.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Rl-Mg3pSDx_GyNxC" /></figure><p>We drew some lines around what our team should fix and around what the developers themselves should fix. In this case, we defined the boundary as such</p><ul><li>Individual tests are under the control of Product Engineering. Therefore, flakey tests and slow test suites should be sped up by the teams who know the code best.</li><li>The machines that run the tests are our responsibility. Therefore, we needed to make sure that CI builds didn’t fail because the machines were set up wrong or were too slow to stand up.</li></ul><p>If you haven’t heard of SLIs before, <a href="https://landing.google.com/sre/workbook/chapters/implementing-slos/">Google has written about them</a>, <a href="https://landing.google.com/sre/sre-book/chapters/service-level-objectives/">a couple of times</a>. The basics are that an <strong>SLI is the measurable ratio of good things to bad things</strong>. It goes down when performance degrades, and it goes up when performance improves. We turned our areas of responsibility into these SLIs:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kM7biWcRoA57AOlx.png" /></figure><p>Note that for stability, we were not measuring the total number of failed builds. We specifically caught error statuses that were not 1 (which implies a test suite failure) or 0 (which isn’t a failure at all). Similarly, we were intentionally not interested in the total time for a build to finish, just the amount of time a build waited to start. This was the distinction we made because we had control over how many machines were idle and ready to pick up new jobs. If the number of machines was too low, we needed to scale smarter.</p><h3>Implementing Service Level Objectives</h3><p>With these SLIs, we negotiated with the customer to define Service Level Objectives (SLOs) for our two themes — speed and stability. Put simply, an SLO is a target value or range for an SLI or <strong>SLI ≥ target.</strong> In essence, we were deciding what percent of the time our SLIs needed to be above our target.</p><p>SLO targets should be below 100% because you won’t have a perfect product. We live in a world with humans, weather, and uncertainty, which means downtime is inevitable, so we should plan for it. By accounting for downtime in the product experience, we also buy learning and development time for ourselves when we can take risks without ruining our SLI.</p><p>After talking with customers, we decided to focus on stability first and settled on these SLOs:</p><ul><li>99.5% of jobs via Buildkite Jobs in the past 30 days should not experience “infra failures”</li><li>95% of the time, the Average Buildkite Job Latency in the past 30 days should be less than 4 minutes</li></ul><p>We gave ourselves one month to experiment. At the end of the month, whether we had reached our goals or not, we planned to sit down with customers again and see if sentiment toward Buildkite stability and speed had increased. Afterall, the point was to make engineers more productive. Short of tracking their time, the best we can do to measure productivity is talk to them.</p><h3>Make Experimentation Easy</h3><p>Now that we knew what we were aiming for, we needed to get real data and make it easy to see progress. At the time, Datadog had just introduced an <a href="https://docs.datadoghq.com/monitors/service_level_objectives/">SLO widget</a>, allowing us to turn the words we were using above into cohesive dashboard pieces.</p><p>The centerpiece of our dashboard was the SLOs themselves. At the time, our speed SLO was green for “the past 7 days” and “the past 30 days”. Our stability SLO was red across the board, as anticipated. Here’s a screenshot of what these widgets look like today, almost a year after the experiment began:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*jIgo0vNuqOkzIlzE" /></figure><p>Further down the board, we added more granular measurements for specific error types. We made these viewable in just about every way possible — a total count of each error type, ratios of individual failure types to the total number of failures, change week over week for each error type, a graph of errors over time, etc.</p><p>Until we interacted with the data, we did not know which view was most helpful. So we set up multiple ways to see the same data, and as we found one useful, we removed the other three. By the end of the experiment, we had a concise and easy-to-understand board that has made debugging easier since the experiment.</p><h3>Experimenting</h3><h4>Change small things first</h4><p>Once we were confident that we were counting everything our SLI deemed “bad”, we were ready to experiment. Some of these “experiments” were low-hanging fruit like upgrading our version of Buildkite, changing environment variables, and removing redundant setup code.</p><h4>Regroup when needed</h4><p>At the start of the experiment we thought we understood how our failures should be grouped. As we continued increasing stability, though, we realized that the biggest bucket of error types could be broken down further. Once we rearranged buckets by identical failure messages, not just the same exit codes, we gained clearer priorities.</p><h4>Roll with the punches</h4><p>One of our changes was purely reactionary. While we were working on another failure cause, a <a href="https://github.com/pry/pry#runtime-invocation">binding.pry</a> slipped into a commit on a feature branch. You may think this should not be a big deal since it didn’t get out into the development or master branches. But since all of the machines for a queue are in the same pool, it tied up machines, caused other jobs on those machines to fail when the machines were marked unhealthy, and drove wait times through the roof. The fix was simple — set a global environment variable (DISABLE_PRY=1).</p><h4>Big swings</h4><p>A bigger change that we wanted to tackle was upgrading our machines from Ubuntu 14 (Trusty) to Ubuntu 18 (Bionic). By the time we were considering this change, we were green for our 7 day and 30 day stability SLO, so we had space to take bigger swings.</p><p>The short story is that this change didn’t work out within a month. Though we didn’t complete this big upgrade within the experiment time, we were able to try and fail because we had SLOs that kept us accountable to our customers. Without that built-in downtime, we might have been too scared to take a big risk and get feedback so quickly. As it was, we were able to identify the scaling and shut down problems quickly.</p><h3>After the experiment</h3><p>The experiment ended with us reaching our goals across the board — all of our SLOs were met for Buildkite’s speed and stability.</p><p>A few months after the experiment, we saw a dip in our stability SLI below the target objective. We didn’t hear a single complaint from engineers about stability in Buildkite while we were investigating the cause. After a week of this, we decided to talk to our customers and decrease our SLO target to 95% instead of 99.5%. We made this decision with the assumption that we had set the bar too high for ourselves in the first place. Other issues were more variable and more intrusive to engineers, so we could ease back without causing undue slowness.</p><p>On a related note, we haven’t changed the speed SLO at all since implementing it. This SLI has only slipped under its target on rare occasions, which were quickly remedied and received little notice from customers.</p><h3>Concluding Thoughts</h3><p>If there’s a place your customers want improvement, start with SLIs and SLOs for your product. This allows you to focus on moving a needle you can see, as opposed to the one that is unmeasurable or just in someone’s head. This also allows you to prioritize issues automatically. For example, if all SLOs are met, the complaint that one build is slow can be put on the back-burner and not distract from regularly-scheduled work.</p><p>If you feel like you’re drowning in maintenance or operational work, embrace SLOs.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=61bb5994d202" width="1" height="1" alt=""><hr><p><a href="https://engineering.gusto.com/service-level-objectives-for-engineering-on-call-peace-of-mind-61bb5994d202">Service Level Objectives for Engineering On-call Peace of Mind</a> was originally published in <a href="https://engineering.gusto.com/">Gusto Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Day in the Life of a Gusto Engineer: New Grad Edition]]></title>
            <link>https://engineering.gusto.com/day-in-the-life-of-a-gusto-engineer-new-grad-edition-d0fb35fbaaa9?source=rss-7b0cecf1231------2</link>
            <guid isPermaLink="false">https://medium.com/p/d0fb35fbaaa9</guid>
            <category><![CDATA[startup-lessons]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[engineering]]></category>
            <dc:creator><![CDATA[Lindsey Whitley]]></dc:creator>
            <pubDate>Tue, 15 Dec 2020 19:21:15 GMT</pubDate>
            <atom:updated>2024-06-25T19:17:52.172Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/479/1*pECBfRubImcosfFfVNZhHg.png" /></figure><p>It’s been two years since we gave the world a look at a <a href="https://medium.com/gusto-engineering/a-day-in-the-life-of-a-gusto-engineer-df95a636329b">“normal day” as a Gusto engineer</a>. Since then the org has grown nearly three-fold and matured as an organization. So we decided to take another crack at it; this time, we’ve talked to engineers from different teams to give a wider view of life at Gusto.</p><p>For this post, we specifically talked to engineers who are about a year out of college and who started working at Gusto right after graduating. They are at a critical stage in their careers, with lots of opportunity and hunger for growth, as well as lots of options about how to accomplish their goals. Olu Gbadebo (Benefits Engineering), <a href="https://www.linkedin.com/in/mariam-ali-issa/">Mariam Issa</a> (Risk Engineering), and <a href="https://www.linkedin.com/in/sofia-carrillo-15bb41136/">Sofia Carrillo</a> (Accountants Engineering) gave us a peek into their work from the daily operations of being a Gustie to how they serve our customers and what sets Gusto apart from the crowd.</p><figure><img alt="Mariam Issa, a member of the Risk Engineering team" src="https://cdn-images-1.medium.com/max/1024/0*1U0wscpUP0EQnB87.jpg" /><figcaption><em>Mariam Issa, a member of the Risk Engineering team</em></figcaption></figure><h3>Impact on Customers</h3><p>Each team has a unique opportunity to make our customers’ lives better. These new grads represent three distinct teams whose work is linked by its impact on the small businesses we serve. In these times when small businesses need us to deliver for them now more than ever, the outcomes of an engineer’s work at Gusto is even more pronounced and more motivating.</p><p>Olu Gbadebo works on a pod within the Benefits Engineering team which, as he describes it “helps small businesses decide the best benefits for them by supporting the Customer Experience (CX) Advising team.” As early as his second week at Gusto, he was able to push code live. He works closely with CX to build new features, incorporate feedback, and fix reported bugs. The highest priority project he worked on made it possible for our customers to see multiple benefits options they could choose from, allowing them to better fit their health benefits to their employees and their business needs.</p><p>Sofia Carrillo’s team builds features to save accountants time, so they can focus on advising and supporting their small business clients. She has spent the past six months improving the reports that accountants use to access their clients’ payroll and benefits information. Most recently, she built a new feature so accountants can download quarterly tax forms in bulk, separated into folders for each client, instead of having to painstakingly download each form individually. This is a huge time savings, particularly for accountants with a lot of clients: in one download, they have all the information they need every quarter. Sofia has also focused on making small UX changes-like making a client table searchable or adding a notification when a report is ready for download-that make a big difference to our accountant partners.</p><p>Mariam Issa works on Gusto’s Risk Engineering team, which focuses on minimizing risk by detecting fraud and proactively preventing credit defaults to protect Gusto and our customers. Mariam helped redesign processes that allow us to offer faster payroll and contractor payment speed for customers. Following the typical ACH pattern, payroll needs to be run 4 days before employees are paid, but these tools help us speed up the process to 1 or 2 days. By increasing the speed of payrolls, we relieve cash flow stresses that small businesses often face. In the past year, she’s also built new tools and automated existing internal tools for Risk Operations, giving our customers faster service.</p><h3>Support</h3><p>As a new grad, having the right support and the safe space to fail, learn, and ask questions can make all the difference. Not only do engineers need to grow their technical skills early on, but they need to be constantly honing their “soft” skills too. All three new grads mentioned other engineers at Gusto as the biggest source of their technical learning and collaboration experience in the past year.</p><p>Pairing as part of the onboarding and daily routine at Gusto help make the transition and learning process easy. Prior to joining Gusto, Olu had never developed with Ruby on Rails, but his team has helped the Ruby on Rails learning journey through pairing and sharing valuable resources. Mariam said, “The engineering community here at Gusto is incredibly supportive. You’ll find more often than not, engineers are willing to offer time to pair to resolve issues if you ask for the help.” Sofia was placed on a team with a lot of experienced engineers, and being the only junior engineer, she had the chance to pair with her teammates and learn from each of them.</p><p>Gusto’s mentorship program, which is open to all engineers, also contributed to Sofia and Mariam’s success. The program matches engineers from different teams and varying experience levels. Instead of being focused on pairing or code reviews, like a teammate, a mentor offers a less technical perspective on career growth by checking in and offering more general guidance biweekly.</p><p>Each engineer’s manager, or People Empowerer (PE), was mentioned as especially valuable to the technical learning process, but also in the transition into full-time working life. “The biggest pillar of support,” Sofia said, “has been my managers, who are trained to focus on goals and growth as opposed to just making sure you just ‘get your work done’. Both of my PEs so far have been supportive and encouraging, while still challenging me to stretch myself and grow!”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*ztGVHqfjUoP5LKng7EP-5g.jpeg" /><figcaption><em>Olu Gbadebo, a recent graduate on Benefits Engineering</em></figcaption></figure><h3>Daily Operations</h3><p>Communication, especially in a distributed team environment, is essential to how our engineering teams operate. Olu’s pod on Benefits Engineering has a 15 minute stand-up every day to talk about what got done the previous day, what <a href="https://docs.github.com/en/free-pro-team@latest/github/collaborating-with-issues-and-pull-requests/about-pull-requests">pull requests (PRs)</a> need review, and what is coming up today. Some teams, like Sofia’s Accountants pod, sometimes meet on Zoom for standup, but mix in asynchronous standups in Slack.</p><p>Aside from standups, we use several tools for communication about what work is being done and other logistics. Engineers at Gusto use email, JIRA, Slack, Zoom, and Github on a daily basis. An engineer can pick up a JIRA ticket to work on and request a pairing session on Slack with other engineers while working on the ticket. Depending on the size of the PR, engineers typically spend less time reviewing than writing code. Most PRs are small and digestible, usually requiring no more than 20 minutes per PR. In some cases, especially for larger pull requests, code reviews and pairing go hand-in-hand helping spread knowledge while speeding up the process.</p><p>All engineers at Gusto spend portions of their day writing code, reviewing code, and pairing. However, for Olu and Mariam, the largest percentage of the day is spent solo coding with about 40% of that time devoted to testing. Sofia says that she spends the most time testing, but solo coding is next in line. All three of them said that pairing with other engineers happens regularly each week, with more pairing happening right when they joined Gusto. Olu added, “Every engineer on the team is always eager to pair.”</p><h3>What sets Gusto apart?</h3><p>As a new grad starting a software engineering career, there are a lot of factors that can differentiate companies and experiences from each other. Some points can be easily measured like compensation, benefits, and location. Others, like the culture of the company and the day-to-day work itself, vary from perspective to perspective.</p><p>Mariam values the fact that she was given the opportunity to work on high impact projects from the get-go. She loves knowing that she’s supported and challenged to consistently grow. Since her manager has made Mariam’s career goals a high priority, she has been able to keep doing impactful work.</p><p>From Olu’s perspective the opportunities at Gusto to learn and grow as an engineer, especially for a new grad, distinguish it from other companies. He also said, “I’ve felt that Gusto is a great place to make an impact. In my first few weeks, I started launching features that helped the CX team.”</p><p>Sofia said that the biggest separator for her was the community. There are a lot of engineering folks who really care about making our engineering community a collaborative and supportive environment. Since this is her first job out of college, she “wanted to make sure [she] was joining a company that was truly invested in growing and mentoring junior engineers.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZkVgkWy_ydL__1nULfZjWg.jpeg" /><figcaption>Sofia Carrillo, an engineer working to improve the accountant experience on Gusto Pro</figcaption></figure><h3>What have you learned?</h3><p>In her first year after graduating, Sofia said that her biggest challenge has been communication and coordination. To quote her directly, “Nothing in school prepares you for working in a large organization with many different departments. Everyone has different knowledge and perspectives that they’re bringing to a meeting, so it’s important to make sure people are communicating and reaching alignment. I’ve learned that it’s hard to strike the right balance of being clear and giving context to everyone who needs it. Whether it’s in person, Slack, documentation, or code, communicating clearly and effectively is harder than you might think!”</p><p>Olu’s biggest learnings have been getting comfortable with Ruby on Rails as his primary framework for writing code. Mariam said, “My biggest learning so far is that there’s not enough time to be hesitant about asking questions.”</p><h3>Transitioning from School to Work</h3><p>Going from college to a career can be scary. Not only are there a lot of <a href="https://academic.oup.com/jxb/article/60/3/712/453685">known unknowns, but there is also a sense that the unknown unknowns</a> are more numerous and more alien than before. Finding the right blindspots can help in making good decisions with confidence.</p><p>Sofia points to the mentality shift that she learned at Gusto around tradeoffs as one of the biggest adjustments for her. “I think school primes you to think there are one or two ‘right’ solutions to a problem. But a lot of times those solutions don’t take into account real world things like deadlines, shipping value to users quicker, and other real life constraints that engineers have to account for.” Good engineering is less about the “right” solution and more about deciding between tradeoffs.</p><p>Olu says working at Gusto has reinforced the value and importance of testing software. “All types of testing should be considered for both big and small projects. This includes unit, integration &amp; QA testing. A lot of bugs can be prevented by writing good tests.” He also emphasizes the importance of a growth mentality for success. “Be quick to learn, be quicker to implement, and be comfortable with finding answers along the way. Your first job out of college is an opportunity to create a learning environment for yourself.”</p><p>Mariam summarized the transition from college to work as such: “The ability to adapt quickly, particularly when working at a fast-paced tech company like Gusto, is essential (especially during a pandemic and economic crisis). Asking questions and being transparent about growth areas earlier on would have served me well since being comfortable in this space enabled me to build the right products and have a larger impact. I also began to recognize that a lot of other engineers share similar experiences early in their careers, so they were able to point me to helpful resources and give insight to their acquired knowledge.”</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d0fb35fbaaa9" width="1" height="1" alt=""><hr><p><a href="https://engineering.gusto.com/day-in-the-life-of-a-gusto-engineer-new-grad-edition-d0fb35fbaaa9">Day in the Life of a Gusto Engineer: New Grad Edition</a> was originally published in <a href="https://engineering.gusto.com/">Gusto Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What is Data Engineering?]]></title>
            <link>https://engineering.gusto.com/what-is-data-engineering-65515c1df9af?source=rss-7b0cecf1231------2</link>
            <guid isPermaLink="false">https://medium.com/p/65515c1df9af</guid>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[data-warehouse]]></category>
            <category><![CDATA[data-lake]]></category>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[getting-started]]></category>
            <dc:creator><![CDATA[Lindsey Whitley]]></dc:creator>
            <pubDate>Thu, 22 Aug 2019 15:30:41 GMT</pubDate>
            <atom:updated>2024-06-25T19:19:55.212Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/949/0*SR_i9r8dk3YZhlkT.jpg" /><figcaption>Swimming in the Data Lake</figcaption></figure><p>Before I started as a Data Engineer two years ago, I had no idea what the role entailed or how it differed from data science and data analytics. Job titles with the word “data” in them are known to be an enigmatic black box. That’s true even for folks in technical roles. This post is what I would have wanted to read when I was trying to fit the pieces of the data pipeline together.</p><h3>Can you give me an overview first?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oGjeteGw-mg5Ki8-d7S7YA.png" /></figure><h3>How do you relate to the other data teams?</h3><p>At Gusto, Data Platform Engineers support the other data teams — Data Analysts and <a href="https://www.sas.com/en_us/insights/analytics/what-is-a-data-scientist.html">Data Scientists</a>. If those teams need access to data, we move and adjust it to a usable format for them. If they need a server that can crunch numbers for them quickly, we maintain that server. If they need a tool to explore data interactively, we make sure they have it — whether it’s an external tool that we keep connected or a tool we build ourselves.</p><h3>What part of the business does your work impact?</h3><p>Primarily, our work impacts the business side of Gusto. For example, we recently switched our customer satisfaction (CSAT) surveys to a new vendor, and the customer-facing teams who rely on the surveys needed to be able to access the data in a similar way as before.</p><p>My job was to collect the raw responses from the third-party vendor, transform the data to make it query-able, and then make sure a partner on the Data Analytics team had access to it. From there the Data Analyst made graphs and tables so that the customer-facing teams were able to see their stats in a way that was already familiar to them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*W2FopBZYJ9syr7hw.jpeg" /><figcaption><em>By supporting Data Scientists, Data Analysts, and anyone who can write SQL, we are able to support the data needs of the whole company.</em></figcaption></figure><h3>Why do you need to do all this to data?</h3><p>If you only needed to use data from one database, you may not need to move it beyond creating a read-only cluster specifically for this purpose. Even then, it’s likely that the database is structured in a way that works for the application or the engineers, rather than for the people who need to analyze the data.</p><p>If your users (i.e. data analysts and business teams) need information from more than one database or want to use data from third party tools, you’ll need some way to group all the data together (i.e. a data warehouse). If your database doesn’t keep a record of each write, you may need some way to allow users to get a historical view of the database.</p><p>I could keep going, but you get the idea. The data is very unlikely to be in the most ideal state for queries that yield fast insights, which is why it needs to be copied and restructured.</p><h3>So, you get data from more than one place? What are all the data sources?</h3><p>In our case, we gather data from our four internal databases which contain the data created by our apps. For example, whenever a customer runs a payroll, multiple rows are created or updated across multiple tables.</p><p>We also gather data from external tools via their APIs. These include Qualtrics, Marketo, and Zendesk Chat, among others.</p><p>Another way to get data from an external tool is to connect to the database directly. We do this with Salesforce, but we also copy data into certain Salesforce tables from our internal databases.</p><h3>Ok, but what product do you create?</h3><p>In our case, we expose a <a href="https://www.forbes.com/sites/bernardmarr/2018/08/27/what-is-a-data-lake-a-super-simple-explanation-for-anyone/#76ca2fa676e0">data lake</a> to the Data Analytics and Data Science teams. The data lake is meant to be a place of discovery for these teams. Since the data is raw, it takes less work for the Data Engineering team to manage, but it doesn’t eliminate data that could be useful for skilled explorers.</p><p>More broadly within Gusto, we expose a data warehouse of tables that are structured to be queried quickly and only contain a subset of all the data in the lake. For Gusto, all of our data goes through the lake before it gets to the warehouse, and only the data that we know is useful and worth cleaning gets to the warehouse. These tables are meant to be more easily understood and allow for varying levels of access to sensitive data through different schemas.</p><p>Overall, we can think about the difference like that between a lake and tap water. Lake water in some places is safe to drink, but usually it’s not and in most cases getting the water is a lot less convenient than drinking from the tap. Tap water on the other hand is enriched with fluoride, is stripped of lead and bacteria, and is at your fingertips when you want it.</p><p>Similarly, the lake is a huge amount of data (all that my team collects). To make any use of it, you need to know what’s in it and how to treat it to write a query that returns anything a human can make sense of. The warehouse has been cleaned for you; it’s in tables that make sense for known use cases; and you can get answers out of it quickly.</p><h3>How do you make the lake and the warehouse?</h3><p>First, we run a scheduled ruby CLI command to get raw data from the data sources and store it in <a href="https://aws.amazon.com/s3/">S3</a> in the rawest possible format. Storing raw means that if we decide later on that we want an additional field or a mapping change, we won’t have to reload everything, we’ll just change the next step.</p><p>Next, we use <a href="https://spark.apache.org/">Apache Spark</a> or <a href="https://aws.amazon.com/athena/">Amazon Athena</a> to perform a simple ETL (Extract<a href="https://www.datawarehouse4u.info/ETL-process.html">, Transform, Load</a>) on the data that we want in the data lake. An ETL could be as simple as reading a table from json files, making small changes to the structure, and writing the new table to csv files. The data lake is also stored in S3, but all files are parquet, whereas the raw data could be collected in any format, like json or csv. For our purposes, this ETL stage leaves the data with the same columns as in raw and switches the file format to <a href="https://parquet.apache.org/">parquet</a>. Changing files to parquet, a compact columnar format, means that <a href="http://prestodb.github.io/">Presto</a> (or, more specifically in our case, Athena) can query them very quickly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Y254tAt2ErtguESt.png" /><figcaption><em>An example of breaking down one row of raw data into separate tables and rows</em></figcaption></figure><p>Finally, once we know what to put into the data warehouse, we run <a href="https://aws.amazon.com/redshift/">Redshift</a> SQL queries to perform another ETL which loads the new arrangement into a table in the data warehouse. This ETL stage does things like combine tables, unnest complex column types like arrays or structs, and remove sensitive fields.</p><h3>How do you decide what is going to be in the Data Warehouse?</h3><p>The short answer is we don’t. For the most part, the Data Analytics team decides what should be available and in what format. Sometimes we are given the requirement and then write the code to move it. Sometimes we build the tools that allow a Data Analyst to build and maintain a table on their own.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZCEm-FrRy7SsqdT4.jpeg" /><figcaption><em>How some of our data sources come together to create the lake and then in turn the warehouse</em></figcaption></figure><h3>I think I followed most of that… is that what all Data Engineers do?</h3><p>Not all Data Engineering teams will do the same work. Some Data Platform teams focus exclusively on infrastructure, while ETL teams might exclusively handle the process of moving and cleaning data.</p><h3>That sounds interesting; how can I learn more?</h3><p>I can’t say exactly what you’ll need to know, but here’s what helped me feel confident in my role:</p><ul><li>Take time to learn more about infrastructure. We use <a href="https://www.docker.com/">Docker</a> to deploy most of the projects that we own, so being able to work with the CLI, read a Dockerfile, and dig into containers has been immensely helpful. I’ve also gotten more comfortable building pipelines in <a href="https://buildkite.com/">Buildkite</a>, writing <a href="https://www.terraform.io/">Terraform</a> files, and navigating the AWS console.</li><li><a href="https://www.w3schools.com/sql/default.asp">Brush up on your SQL</a>. You might not be writing a ton of raw SQL right now, but as a Data Engineer you’ll likely being writing more complicated SQL and need to be able to understand the queries that other data teammates will be writing.</li><li>Ask lots of questions and be fearless. There are still projects that my team owns but that I haven’t worked on a ton. I push myself to pick up tickets in those domains with the knowledge that I can ask my teammates questions, even though it is usually scary to work on new code and admit what I don’t know.</li><li>Keep looking for <a href="https://www.analyticsvidhya.com/blog/2018/11/data-engineer-comprehensive-list-resources-get-started/">more resources</a>.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=65515c1df9af" width="1" height="1" alt=""><hr><p><a href="https://engineering.gusto.com/what-is-data-engineering-65515c1df9af">What is Data Engineering?</a> was originally published in <a href="https://engineering.gusto.com/">Gusto Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[When Girls Lose Interest in STEM — and What You Can Do about It]]></title>
            <link>https://engineering.gusto.com/this-is-when-girls-lose-interest-in-computer-science-and-heres-what-you-can-do-about-it-a1bbdf394978?source=rss-7b0cecf1231------2</link>
            <guid isPermaLink="false">https://medium.com/p/a1bbdf394978</guid>
            <category><![CDATA[women-in-tech]]></category>
            <category><![CDATA[girls-in-tech]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[girls-in-stem]]></category>
            <category><![CDATA[engineering]]></category>
            <dc:creator><![CDATA[Lindsey Whitley]]></dc:creator>
            <pubDate>Fri, 08 Mar 2019 01:14:28 GMT</pubDate>
            <atom:updated>2024-06-25T19:07:05.532Z</atom:updated>
            <content:encoded><![CDATA[<h3>When Girls Lose Interest in STEM — and What You Can Do about It</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*zHVeVmqF_RUlznSQ.jpg" /><figcaption><em>Patches our design team created for the event</em></figcaption></figure><p>I’m an engineer at <a href="https://gusto.com/">Gusto</a>. In my spare time, I teach middle- and high school-aged <a href="https://medium.com/gusto-engineering/what-i-learned-by-teaching-fourth-grade-girls-to-code-1d678fb5150c">girls how to code</a>. I do this because I want girls to know about all the career options that are available, even if their communities don’t.</p><p>As I was researching the best ways to get students interested in STEM fields, this sentence stood out:</p><p><a href="https://money.cnn.com/2017/02/28/technology/girls-math-science-engineering/index.html">“Girls become interested in so-called STEM subjects around the age of 11 and then quickly lose interest when they’re 15.”</a></p><p>Up until this point I had only focused on getting younger (i.e., elementary school) girls interested in computer science, hoping that the spark would last through college. But now I realized that I was missing a huge part of the equation.</p><h3>Why girls lose interest in STEM starting in high school</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1w8egqu992oWfhKx.jpg" /><figcaption><em>Photo by Ben Mullins on Unsplash</em></figcaption></figure><p>High school is angsty.</p><p>I remember worrying about being liked by my friends and wanting to be noticed by a boy, all of which made it harder to focus on the future and who I wanted to be.</p><p>Had it not been for my desire to rebel against what society told me I couldn’t accomplish, I don’t know whether I would have continued to focus on STEM.</p><p>Today, most of my friends who are in STEM fields became interested in it after college. Welcoming people without formal training into roles they are capable of may be more common today, but it is still easier to prove to a prospective employer that you’re worth hiring when you have formal training and start earlier.</p><blockquote><em>This means that keeping young women interested in STEM — well beyond middle school — opens up more career opportunities for them earlier in life.</em></blockquote><h3>What we’re doing about it</h3><p>While developing a program to keep girls interested in STEM, I was inspired by shadowing sessions we did with a <a href="https://turing.io/">Denver code school called Turing</a>. In these sessions, post-high school web development students came to the Gusto office, took a tour, asked questions about working here, and then worked with an engineer for an hour.</p><p>Now that we were targeting high school students, who are younger, we needed to adjust our tactics. So here’s how we kicked off the program.</p><ul><li><strong>We partnered with </strong><a href="https://www.girlscoutsofcolorado.org/"><strong>Girl Scouts of Colorado</strong></a>, which has a large pool of engaged high school students. They set up an application for our event and included an announcement in a newsletter they send to high school Girl Scouts. Creating an application helped us find girls who were genuinely interested in the event and willing to put in a bit of effort to be part of it.</li><li><strong>We made the experience worth students’ time.</strong> Since we anticipated that girls would be driving into Denver from the suburbs, we wanted to make each session longer (about four hours). This would give them a better travel-to-activity ratio.</li><li><strong>We targeted all levels of experience.</strong> We assumed that some girls would have written code before and some probably wouldn’t know the basic concepts of coding.</li><li><strong>We went beyond just engineering.</strong> Unlike the Turing students who were further along in their STEM-journeys, we couldn’t assume that these younger girls had decided to do anything related to STEM. To account for this, we expanded our scope to include Product and Design.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HLMo-jG9I0IVqhRnwyIm2Q.jpeg" /><figcaption><em>Photo by STEMShare NSW on Unsplash</em></figcaption></figure><h3>How we announced the program</h3><p>I sat down with teacher-turned-engineer <a href="https://www.linkedin.com/in/stephanie-bentley/">Stephanie Bentley</a>, to go through the <a href="https://www.cde.state.co.us/standardsandinstruction/computer-science-cas-spring-2018-committee-recommendations">Colorado Academic Standards for Computer Science</a> and draw connections between what we planned to do and what the state expects students to learn about CS.</p><p>For the announcement and application, our blurb looked like this:</p><p><em>Girl Scouts is partnering with Gusto, a Silicon Valley- and Denver-based technology company, to teach girls entering 10th, 11th, and 12th grades what it’s like to write code that solves a real-world problem! Gusto will host a group of girls for three (on November 16th, April 12th, and June 28th) four-hour (from 10am-2pm) “jam with software developers, product managers, and designers” sessions with lunch included. In each session, girls will get a taste of an average day as a developer — from creating a feature to finding and fixing bugs in the code. Over the three sessions, we’ll go through a design sprint to define a problem to tackle with software and design a solution. We hope to inspire a future cohort of computer scientists!</em></p><p>Missing school can be hard to justify, which is why we made sure the program addressed the state’s computer science standards:</p><p><em>1.4 Large, complex problems can be broken down into smaller, more manageable components</em></p><p><em>2.3 Computer software is written for specific purposes</em></p><p><em>2.5 Client considerations drive system design</em></p><p><em>3.1 The creation of a computer program requires a design process</em></p><p><em>3.3 Collaborative tools, methods, and strategies can be used to design, develop, and update computational artifacts</em></p><p><em>3.4 Client-based design requirements and feedback are essential to a quality computational product or service</em></p><p>Planning beforehand only required a few hours a week for a few weeks, including a couple of meetings with Stephanie to make sure we weren’t totally missing the mark.</p><h3>The day of</h3><p>On the day of the event, I met the girls and their parents, and signed them in. Then we all got on the same page by discussing:</p><ul><li>Our goals for them</li><li>Our expectations for their participation</li><li>Who we are: What are the teams here? What is the structure? How do Engineering, Product, and Design work together?</li><li>What do they want out of this? What’s their background? How familiar they actually are with code? (Depending on what they knew, we could target who they were shadowing.)</li></ul><p>During the design sprint, we explained why we do design sprints and then what our goal for the day was — to pick a problem to tackle in the coming sessions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZQbKxiTct-YhDxJk.jpg" /></figure><p>We started by brainstorming. In ten minutes, each of us got a stack of post-its and a pen and wrote as many problems that we encounter in our daily lives as we could think of. We then explained each idea to the group and tried to group problems together as we went. Then the team voted on which problem we wanted to tackle together.</p><p>Once we had a winner, we spent the rest of our time digging into what problem we really wanted to solve and which angle was most interesting to all of us.</p><p>At the end of our session, we landed on the problem: Teenagers and adults have a hard time getting past discussions about the weather and into a deeper conversation. So how can we fix that?</p><p>Since the group was pretty evenly split between teenagers and adults, we felt particularly able to explore and solve this problem.</p><p>Talking about the generational gap is valuable on both sides: adults can share their school and work experiences, and teenagers provide fresh insights. In the design sprint sessions to come, we will nail down the root cause of this gap and start designing a solution to the problem and define an MVP.</p><h3>High school kids are smarter than you think</h3><p>One of my concerns at the start of the day was that we were throwing too much at these students in such a short period of time. However, no one was too shy to interrupt, ask for clarification, or express particular interest in a concept.</p><p>While explaining a tool that my team manages, how we deploy, and what happens as a result of deployment to one of the students, I was grateful that I had jumped right in and allowed her to pick our direction by asking good questions.</p><p>In our debrief at the end of the day, we asked everyone:</p><ul><li>What was the most valuable part of the day?</li><li>What could you have done without?</li><li>Do you have any general feedback on how everything went?</li></ul><p>We learned that while it’s still valuable to ask participants what they thought in-person, there might be a better way to do this. Maybe we could have asked throughout the day, for rankings 1–10 for different activities, or in a way that gave them anonymity. This is something we’ll be changing before the next two sessions.</p><p>Re-reading the articles that led to this program, I remembered that girls often have misconceptions about STEM. When we ask for feedback in the next two sessions, I’ll focus on understanding whether we are addressing these common issues:</p><h3>What we hope to do next</h3><p>In the next five months, we’ll have two more sessions like this where we finish our design sprint and continue to demonstrate the social, logical, and transformative nature of writing software. Designing and creating a solution to a problem transforms it from a nebulous annoyance into a functional product.</p><p>By pairing with the girls, eating lunch together, and allowing them to observe engineers working together, we hope they see that good code is written when people feel comfortable to bounce ideas off each other and pool their collective knowledge.</p><p>I’m excited to see what we learn about why girls do or do not pursue STEM — and whether our program can make a difference.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a1bbdf394978" width="1" height="1" alt=""><hr><p><a href="https://engineering.gusto.com/this-is-when-girls-lose-interest-in-computer-science-and-heres-what-you-can-do-about-it-a1bbdf394978">When Girls Lose Interest in STEM — and What You Can Do about It</a> was originally published in <a href="https://engineering.gusto.com/">Gusto Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What I Learned By Teaching Fourth Grade Girls to Code]]></title>
            <link>https://engineering.gusto.com/what-i-learned-by-teaching-fourth-grade-girls-to-code-1d678fb5150c?source=rss-7b0cecf1231------2</link>
            <guid isPermaLink="false">https://medium.com/p/1d678fb5150c</guid>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Lindsey Whitley]]></dc:creator>
            <pubDate>Tue, 19 Jun 2018 18:58:03 GMT</pubDate>
            <atom:updated>2024-06-25T18:58:34.508Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*V4ejNbw8SnTSza3t.JPG" /><figcaption><em>Emily helping a student get started.</em></figcaption></figure><h3>How it started</h3><p>A teammate and I were headed downstairs for lunch when he said something like, “Beyond focusing just on Gusto, we should be doing more to change the ratio of women to men in engineering.” Little did he know I had been toying with the idea of teaching girls to code, but I wasn’t sure until then that I wanted to do it with Gusto.</p><p>Choice is power. Yes, I love my coworkers, the challenges of data, the ever-changing landscape of software, and the chance to solve problems for our customers. Ultimately, though, I’m grateful for my job because it gives me the means to be independent. This industry provides flexibility to learn new things every day, to work with different teams, to live almost anywhere, to adjust my schedule for bumps in the road, or to save/invest/spend as I please. I am lucky that I grew up with a family and community that always encouraged my agency and voice. Teaching girls to code empowers them in the same way by exercising their brains and pushing their boundaries.</p><p>Since I hadn’t done anything like this before, I talked to our Head of Engineering in Denver to pick his brain about any red tape. Gusto was totally supportive, so all I had to do was plan and execute the event.</p><h3>What happened</h3><h4>1. Garner support from my team.</h4><p>I wasn’t about to teach fifteen to twenty girls to code by myself, so I needed to enlist the help of my fellow engineers. To get them involved, all I really had to do was explain the vision I had for this event — getting girls to identify with “engineering” as a viable career option — and they were in. Their involvement ended up being key to the quick execution of this event because ten heads are way better than one.</p><p>I talked to a few of my teammates individually first; then I brought it up with the whole team at lunch. Simply saying, “Let’s teach elementary school girls to code. I’ll figure out a lesson plan and organize the event. Will you help facilitate?” was enough to whet their interest. The only further explanation needed was that the weakness of diversity in our pipeline is due, in part, to the overall lack of diversity in the industry, and we can start changing that more proactively.</p><h4>2. Find girls to teach.</h4><p>Our goal was to teach girls from a mix of socio-economic and ethnic backgrounds. Dora Moore teaches a mix of students that reflects the demographics of the diverse CapHill neighborhood of Denver.</p><p>A stroke of luck — a teammate’s friend, Stephanie, is a fourth-grade teacher at a school about two miles from our office, called Dora Moore School. Even better, when we talked to Stephanie, she was as excited about this idea as we were. She brought in the other fourth-grade teacher, Dory, which allowed us to teach every single fourth-grade girl at the school.</p><h4>3. Decide what to teach.</h4><p>Since I don’t have much experience creating curricula, especially for fourth graders, I searched online for what to teach. I found an organization called <a href="http://www.techgirlz.org/techshopz/">TechGirlz</a> that provided different lesson plans and platforms, including the one we used, called <a href="https://blockly-games.appspot.com/">Blockly</a>. Blockly has a few different “games,” of which we used the <a href="https://blockly-games.appspot.com/puzzle?lang=en">Puzzle</a> and the <a href="https://blockly-games.appspot.com/maze?lang=en">Maze</a>. To make sure that all of the students understood the basics, we started them out on the Puzzle. Afterward, the Maze introduced loops and if/else statements while giving them a way to relate to the code spatially.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/828/0*BKhC78LWFaRvmiDG.png" /><figcaption><em>An example problem used in the workshop with its solution. In this case, the number of blocks (shown on the right) are limited to avoid brute force solutions.</em></figcaption></figure><h4>4. Organize logistics.</h4><p>I worked with Stephanie to arrange for us to come to her classroom. This was easier than having the students come to Gusto because it avoided permission slips from parents and the transportation of twenty students. She handled the school logistics and I managed to find a two-hour window for seven engineers to help. Fortunately, the classroom had enough Chromebooks for the students so we didn’t have to come up with twenty laptops to take with us.</p><h4>5. Prep the instructors.</h4><p>Another stroke of luck — Emily, one of our San Francisco-based engineers, was in Denver for a three-month rotation. Aside from being an extremely capable engineer, she has experience working with kids and the confidence and charisma to keep order in a classroom. With Emily on board to teach, all I had left to do was finalize logistics.</p><p>After I created the slide deck, the team convened to go over the plan and the activity for the event. This was also a great time to get feedback on the presentation, the plan for timing, and the activity itself.</p><h4>6. Do the workshop.</h4><p>We got to the classroom about fifteen minutes before we planned to begin and got set up. Emily led the intro discussion about what engineering is, why she does it, and what we’d be doing. Then we were off to the races. We wrapped up about ten minutes early to give the girls time to ask Emily, Kendra, and me any questions they had.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wl2N7wYLKX_IHBV9.JPG" /><figcaption><em>The team setting up laptops and checking that the tool would be accessible on the school’s wifi.</em></figcaption></figure><h3>What we learned</h3><h4>Before deciding what curriculum to use, understand the students’ current knowledge.</h4><p>The fourth graders we taught hadn’t taken geometry yet. This kind of information is probably accessible online, but with such a helpful teacher, I could have asked more questions to understand how to incorporate the math they are learning into the activity.</p><p>Unbeknownst to me, the girls have a technology class and had done projects similar to Blockly. If I had known this, we could have spent a bit less time on the intro and more time on more interesting problems.</p><h4>There will always be a wide range of prior knowledge among students.</h4><p>For example, one girl finished the project on her own, but another needed much more help to keep moving along. Combat this disparity by using part of the workshop to do pair programming. The first time around, pairing was an afterthought, and the girls received no instruction on effective techniques. Giving ground rules for how to pair and setting the expectation that it would be part of the activity would have encouraged teamwork and avoided diminishing self-esteem brought on by asking for help.</p><h4>Provide focus for questions.</h4><p>Fourth graders are curious about everything. If you don’t provide them with guidelines, questions can quickly go from “How did you know you wanted to do this?” to “How old are you?” If you do remind them what you’re there for, they will ask insightful, helpful questions.</p><p>Topics we might use to focus their questions in the future include: how we collaborate, what educational path we took to get where we are, and what an average day consists of.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*QH9ncqsvEdlvcDDS.JPG" /><figcaption><em>Lindsey and Emily answering questions from the class.</em></figcaption></figure><h4>Sell the idea that coding is a valuable skill.</h4><p>To lead the companies of the future, they need to know about technology, even if they don’t want to write code themselves. They are probably <a href="http://money.cnn.com/2017/02/28/technology/girls-math-science-engineering/index.html">already interested</a>, so use that to your advantage. <a href="https://www.inc.com/rebekah-iliff/5-ways-to-get-girls-to-fall-in-love-with-stem.html">Connect coding to other things they are learning</a> — math, critical thinking, problem solving, etc. Having a basic understanding of software will give them leverage in working with engineers. If they want to be members of Congress, knowing how the software that they will use and oversee works will only make them more successful.</p><h4>Help them learn to fail and persevere.</h4><p>Among the girls who told us that they didn’t like the activity, most were simply frustrated. After doing a bit of research, I found that <a href="https://www.computing.co.uk/ctg/opinion/3031171/to-see-more-girls-in-stem-we-must-boost-their-confidence-not-interest">lack of confidence</a> and <a href="https://www.huffingtonpost.ca/2017/08/17/girls-stem_a_23078434/">fear of failure</a> are two leading reasons that girls don’t consider STEM. If we can help them face and overcome difficulty, we can help build their confidence in their ability to learn.</p><p>This helps them far beyond coding or any academic setting. Whether they hate science or dream of circuits, building strong, resilient, hard-working girls will help them succeed in anything that comes their way.</p><h3>How we’re following up</h3><p>I plan to spend a bit of time in the next few weeks scouring the internet and any academic papers I can get my hands on for the most effective strategies. Some questions that I would like to answer:</p><ul><li>Do girls need to see women in STEM as role models consistently to have an impact? Or does it just take a couple of interactions to make STEM a possibility?</li><li>Are there strategies that we’re missing? Should we do something other than work on small problems?</li><li>Would it be more effective to work on longer projects? Would an on-site at Gusto add to their excitement?</li></ul><p>My intuition is that prolonged presence of role models makes a bigger difference, so I’m reaching out to Dora Moore’s tech teacher to learn more about where we can be involved there.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*eP2iYSK_tcQda1Lq.JPG" /><figcaption><em>Emily and Chris ready to take questions.</em></figcaption></figure><h3>Why I’ll continue</h3><p>I can’t imagine many places where an engineer would be encouraged to use their expensive time to plan events that use even more engineering time to teach girls to code. Not by accident, I have found exactly that at Gusto.</p><p>One day when I was starting to lose steam on follow-up, a teammate reminded me that I cared about the mission of these workshops and that I had the support of my team to accomplish it. With his reminder, I turned back to my keyboard with renewed intention to research, plan events, and, hopefully, inspire others to do likewise.</p><p>For me this chance to nudge my industry toward greater equality is about <a href="https://www.huffingtonpost.ca/2017/08/17/girls-stem_a_23078434/">doing what’s right</a>, more than what’s right for the bottom line, <a href="https://hbr.org/2016/02/study-firms-with-more-women-in-the-c-suite-are-more-profitable">though it accomplishes that too</a>. To tell the truth, my initial thought process was solely focused on <a href="https://edscoop.com/microsoft-stem-women-girls-research-annie-arnold">giving other girls the confidence to do what they want</a>, as my parents and community gave me at their age. The fact that my community still values the same things helps me push harder.</p><p>If you’re interested in not only doing great engineering work, but also contributing to the community, that’s what we do here at Gusto. We’re hiring for engineers in both Denver and San Francisco! <a href="https://gusto.com/careers">Check out our job openings</a>.</p><h3>Resources</h3><h4>Articles</h4><ul><li><a href="https://hbr.org/2016/02/study-firms-with-more-women-in-the-c-suite-are-more-profitable">Study: Firms with More Women in the C-Suite Are More Profitable</a></li><li><a href="https://www.huffingtonpost.ca/2017/08/17/girls-stem_a_23078434/">How To Encourage Girls To Get Interested In STEM</a></li></ul><h4>Tools</h4><ul><li><a href="https://blockly-games.appspot.com/">Blockly</a> — learning platform with easy to use games</li><li><a href="http://www.techgirlz.org/techshopz/">TechGirlz</a> — lesson plans and platforms, help with publicity and overall structure</li><li><a href="https://docs.google.com/a/gusto.com/presentation/d/e/2PACX-1vSZYaO8oT_T_x0ExzYKw3CR6LggCk7ET4UDZVNBqg2GuYzdJQF-caPKdOAmodeNALATtT-PR-uDVQi-/pub?start=false&amp;loop=false&amp;delayms=3000">Slide deck</a> — What we used to teach the class</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1d678fb5150c" width="1" height="1" alt=""><hr><p><a href="https://engineering.gusto.com/what-i-learned-by-teaching-fourth-grade-girls-to-code-1d678fb5150c">What I Learned By Teaching Fourth Grade Girls to Code</a> was originally published in <a href="https://engineering.gusto.com/">Gusto Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>