<?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[Spektrix Engineering - Medium]]></title>
        <description><![CDATA[Meet the engineering and product teams transforming the way arts organisations sell tickets and manage their audience relationships - Medium]]></description>
        <link>https://medium.com/spektrix-engineering?source=rss----f9f911ef2ee---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Spektrix Engineering - Medium</title>
            <link>https://medium.com/spektrix-engineering?source=rss----f9f911ef2ee---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 13:58:29 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/spektrix-engineering" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Brighter and Azure Service Bus — Contributing to Open Source]]></title>
            <link>https://medium.com/spektrix-engineering/brighter-and-azure-service-bus-contributing-to-open-source-7dfa364137d6?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/7dfa364137d6</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Damon Siusta]]></dc:creator>
            <pubDate>Wed, 31 Mar 2021 16:30:52 GMT</pubDate>
            <atom:updated>2021-03-31T16:30:52.196Z</atom:updated>
            <content:encoded><![CDATA[<h3>Brighter and Azure Service Bus — Contributing to Open Source</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e2Odga58h5KRsrZD4M-YpA.jpeg" /></figure><p>Part of the Spektrix migration to Azure involved a significant piece of work to replace our existing on-premise <a href="https://www.rabbitmq.com/">RabbitMQ </a>servers with the more cloud native Platform as a Service (PaaS) <a href="https://azure.microsoft.com/en-gb/services/service-bus/">Azure Service Bus</a> offering. Fundamentally the same principles of message queuing applied (albeit with slightly different terminology) so we were confident we could switch out the service in the components that needed them. In our attempts to begin the process of decoupling our Monolith, we use a <a href="https://martinfowler.com/bliki/CQRS.html">CQRS</a> pattern and to support this we use an open source component called <a href="https://www.goparamore.io/">Brighter</a>.</p><p>Brighter describes itself as:</p><p>“<em>…a Command Processor &amp; Dispatcher implementation with support for task queues that can be used as a lightweight library.</em></p><p><em>It can be used for implementing </em><a href="http://alistair.cockburn.us/Hexagonal+architecture"><em>Ports and Adapters</em></a><em> and </em><a href="https://cqrs.files.wordpress.com/2010/11/cqrs_documents.pdf"><em>CQRS (PDF)</em></a><em> architectural styles in .NET.</em></p><p><em>It can also be used in microservices architectures for decoupled communication between the services</em>”</p><p>In some of our components, we hand off to Brighter to send and receive messages to a message bus and we already had this configured to use RabbitMQ. When we began our analysis of migrating this to Service Bus we realised we had a little bit of a problem. At the time Brighter only had partially implemented support for Azure Service Bus and this wouldn’t be sufficient for our use cases.</p><p>So we got to work writing the implementation and after a few iterations we did it. Of course there were a few challenges along the way — for example, making sure we correctly handled asynchronous methods and events, disposed of connections when we should and that we implemented the package in the way Brighter expected components to behave. We were careful to add the right amount of logging so that we could see how the component was working and adjusting where required. We made sure to handle exceptions, retry transient connection failures with <a href="http://www.thepollyproject.org/">Polly</a> policies and where necessary throw the types of exceptions that the core Brighter component expected so it would function in the way it was expected to.</p><p>We work in an agile iterative manner at Spektrix and once our confidence was high enough we swapped out our RabbitMQ implementation with our Service Bus implementation and our new component performed really well, perhaps just needing the occasional tweak. The last time any significant work was done to this component was around September 2020 and all components needing this integration were modified and successfully deployed to Production.</p><p>Fast forward to March 2020 and some members of the original implementation team were tagged on a Twitter thread:</p><h3>Ian Cooper 💙 on Twitter: &quot;If you are interest in contributing to @BrighterCommand we are very interested in working with a team using Azure Service Bus to get a reliable transport for that platform. None of the core maintainers are on ASB so we need to work with a team who are. / Twitter&quot;</h3><p>If you are interest in contributing to @BrighterCommand we are very interested in working with a team using Azure Service Bus to get a reliable transport for that platform. None of the core maintainers are on ASB so we need to work with a team who are.</p><p>Our previous Principal Engineer, Domenico Musto, had submitted an early pull request but this was before the team had fully completed getting the component into production. Dom tagged us in the thread and asked if we could submit what we had done back to the core Brighter project. We spoke about it at our early morning stand-up and decided it would be best if we ran this past senior management as it would be the first time Spektrix had done something like this. There was enthusiastic support from our Principal Engineer Andrew Jones-Weiss and CEO Michael Nabarro as the opportunity to share our knowledge aligned with the company’s wider principles of collaboration.</p><p>The team put the pull request together by making a few changes to the project to align it with the core Brighter solution (for example namespace changes and a largely painless update from <a href="https://nunit.org/">nUnit</a> to <a href="https://xunit.net/">xUnit</a>). The <a href="https://github.com/BrighterCommand/Brighter/pull/1438">PR</a> was submitted later that day much to the delight of the Brighter core maintainers and other contributors interested in taking this implementation forward. Additional features that we hadn’t added as part of our implementation were required along with some other changes necessary to make the component fully generic.</p><h3>Ian Cooper 💙 on Twitter: &quot;If you build it, they will come https://t.co/dOnAM3NjVk pic.twitter.com/2s3b7tKzGY / Twitter&quot;</h3><p>If you build it, they will come https://t.co/dOnAM3NjVk pic.twitter.com/2s3b7tKzGY</p><p>From the team’s perspective it felt so great to be in a position where we could actively contribute back to the open source community and fulfil an immediate need of members of that community. It’s such a great example of giving back and being recognised for the contribution. I’ve personally worked in companies where this just wouldn’t be possible or the amount of approvals required would just be overwhelming. It’s a true testament to the agility and trust developers are given at Spektrix.</p><p>And as for the future of our bespoke Service Bus component implementation, I can imagine we’ll look to swap out our implementation with the one soon to be provided natively with Brighter and once again reap the benefits of open source software.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7dfa364137d6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/brighter-and-azure-service-bus-contributing-to-open-source-7dfa364137d6">Brighter and Azure Service Bus — Contributing to Open Source</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Box Office App — A Serverless Journey]]></title>
            <link>https://medium.com/spektrix-engineering/box-office-app-a-serverless-journey-942d97bf6380?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/942d97bf6380</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[serverless]]></category>
            <dc:creator><![CDATA[Damon Siusta]]></dc:creator>
            <pubDate>Fri, 19 Feb 2021 15:42:02 GMT</pubDate>
            <atom:updated>2021-02-19T15:42:02.656Z</atom:updated>
            <content:encoded><![CDATA[<h3>Box Office App — A Serverless Journey</h3><p>Over the past two years, Spektrix has been in the process of migrating its services and infrastructure to the Azure cloud. The journey to this point has been fraught with challenges and we are now at the point where we will be able to decommission our own hardware and complete our move to the cloud. The Spektrix platform is now more prepared than ever to support our users as they scale.</p><p>Along the way, we have learnt an incredible amount about Azure cloud services and I thought it would be great to share one of our many success stories: how we took a legacy self-hosted service and migrated it to a cloud native event-driven microservice.</p><p>One of the many services that Spektrix provides to our users is the ability to perform several functions that require integration with devices used in their box offices, things like printing tickets, taking payments from chip and pin terminals and opening cash drawers. As part of the Spektrix solution, we ask users to install a Windows application on their desktops and configure it to suit their requirements (for example the type of ticket printer or the particular provider they make use of for chip &amp; pin payments). The solution itself comes with several key requirements:</p><ul><li>We need to know at any point which applications are actively connected. For example, when printing a ticket from our Spektrix Web application, we need to know which printers are available to perform the printing.</li><li>We need to be able to perform automatic upgrades so we can deliver features and improvements to users seamlessly.</li><li>It must be responsive and reliable. For example, we do not want the customers of our users to wait unnecessarily while their tickets are being printed.</li></ul><figure><img alt="A pair of tickets being handed over at a box office" src="https://cdn-images-1.medium.com/max/480/1*YvJJyqgLO7nJw1aPrticfA.jpeg" /></figure><h3>Core Goals</h3><ul><li>We wanted as much of the solution as possible to use Platform as a Service (PaaS) offerings. At Spektrix, we want to spend our time focusing on solving our users’ needs rather than on maintaining infrastructure.</li><li>We wanted the solution to be scalable, reliable, and secure.</li><li>Communication between the user applications and the service must be fast — for example, the delay between a user requesting a ticket to be printed and the ticket beginning to be printed should be as responsive as possible.</li><li>We need to know as quickly as possible if an app is available, our users should not be able to select a printer connected to an app if the app is not available.</li><li>The auto-upgrade of the user apps should be reliable. We wanted all our users to benefit from having the latest version installed.</li><li>We wanted the solution to be as independent as possible from our other services, this aligns with our strategy of breaking apart our monolith into separate deployable business capabilities.</li><li>It must be cost effective. We wanted to make sure we get the most value out of any service that we deploy to the cloud.</li></ul><h3>Design Challenges</h3><p>During our design sessions we acknowledged that we had several challenges that we needed to overcome before the solution would work:</p><ul><li>Due to historic reasons, we had multiple major and minor versions of the app deployed to our users. Some users had a mixture of major and minor versions also. Version 1 of the application made use of a proprietary web socket technology for maintaining connections to the user. Version 2 of the application made use of a long-polling technique which was not scalable as it impacted the web servers on which we host the Spektrix platform. We needed to ensure both versions supported upgrading to the new version.</li><li>Our strategy of preferring PaaS over Infrastructure as a Service (IaaS) meant that we had taken a decision to move from RabbitMQ to Azure Service Bus. RabbitMQ was used by version 2 of the app so it was a requirement to replace this.</li><li>A lot of this technology was new to us, we needed to quickly iterate and validate that our technology choices were correct and cost effective.</li></ul><figure><img alt="An illustration showing the various Azure components used in the solution along with the interaction with the Spektrix Web Platform and Client Box Office App installations" src="https://cdn-images-1.medium.com/max/1024/1*Xj_-RfB5HapHPBzl_LADew.jpeg" /><figcaption>Architectural High-Level Overview</figcaption></figure><h3>Technology Choices</h3><h3>Azure SignalR Service</h3><p>The <a href="https://azure.microsoft.com/en-gb/services/signalr-service/">Azure SignalR Service</a> gives us the ability to maintain a real-time connection between the Spektrix platform and our user installations of the app. This effectively delegates all the complications about having to concern ourselves with connection management. One of the most beneficial aspects of SignalR is that it provides the ability to auto-recover lost connections. So, if our users are having connectivity issues, the library detects this and attempts to re-establish the connection. As this library also abstracts the underlying connection transport types (WebSocket by default falling back to older transports like Ajax long polling when required) we will be able to benefit in the future if any newer transport types become available.</p><h3>Azure Functions</h3><p><a href="https://azure.microsoft.com/en-gb/services/functions/">Azure functions</a> are a serverless offering from Azure that natively support an event-driven architecture. Functions are triggered based on some external event (such as a new message on the Service Bus or perhaps an HTTP request to the function). They also give us the ability to scale on demand — as more users use the service, Azure will automatically scale the underlying servers hosting our function. This is provided by default by the consumption plan of Azure App Services. As we only pay for when the functions are being executed, in times of low demand, our costs can drop. At the time of writing, the average daily cost in Production for the entire service is around £5, and with some small changes to our Cosmos configuration we are confident we could get this even lower.</p><h3>Azure Cosmos DB</h3><p>We have a requirement to store some state about the user applications connecting to Spektrix, such as when they last connected, the version they are currently running and the descriptive name of the app they have registered. <a href="https://azure.microsoft.com/en-gb/services/cosmos-db/">Azure Cosmos DB</a> also supported our technical requirements of being fast to respond, serverless (in that we do not manage the service), extremely scalable (far more than we would ever need for this use case) and cost effective.</p><h3>Azure Service Bus</h3><p><a href="https://azure.microsoft.com/en-gb/services/service-bus/">Azure Service Bus</a> serves as our message broker and effectively our integration layer between the Spektrix Web platform and the Box Office Service. When users request to perform an action through the Spektrix platform (such as printing a series of tickets), a new message is placed on the bus which an Azure function then consumes sending the necessary message, via SignalR, to the correct app configured printer.</p><h3>Application Insights, Log Analytics and Grafana</h3><p>Monitoring and logging of any production application is essential, and we took advantage of the ‘out-of-the-box’ integration that Application Insights (part of <a href="https://azure.microsoft.com/en-gb/services/monitor/">Azure Monitor</a>) gives us with Azure Functions. We can watch telemetry in real time or query historical telemetry data. We have metrics around performance of each of our functions and smart alerts that fire when anything out of the ordinary occurs. Logs are streamed to Log Analytics where we have created several pre-prepared Kusto queries to enable quick troubleshooting if one of our users is having problems which our support team cannot assist them with. All this data is then visualised via several dedicated <a href="https://grafana.com/">Grafana</a> dashboards which give us a rapid way of monitoring and assessing how the whole system is functioning.</p><h3>Azure Event Grid</h3><p><a href="https://azure.microsoft.com/en-gb/services/event-grid/">Azure Event Grid</a> gives us a great way to integrate Azure services together and we have taken advantage of this by subscribing to user disconnection events from the Azure SignalR service. As soon as a user app disconnects, SignalR produces an event that is published to a topic on Azure Event Grid. We have a function that consumes this event and performs the necessary action to mark that user app as unavailable.</p><h3>Summary</h3><p>Our users are now all running on the most up-to-date version of the box office application, reliably connecting to the Spektrix platform via our new Box Office Service. We have a scalable and easy to maintain platform with no downtime during deployments. We have visibility of the performance of all aspects of the application giving us an improved ability to diagnose and mitigate issues as they arise. We are in a great place going forward to iterate on this design and push more improvements directly into our users’ box offices.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=942d97bf6380" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/box-office-app-a-serverless-journey-942d97bf6380">Box Office App — A Serverless Journey</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Are once sought after standard academic routes into tech still relevant for modern companies?]]></title>
            <link>https://medium.com/spektrix-engineering/are-once-sought-after-standard-academic-routes-into-tech-still-relevant-for-modern-companies-b20ce8137c6b?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/b20ce8137c6b</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[technical]]></category>
            <category><![CDATA[recruitment]]></category>
            <category><![CDATA[diversity]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Emily Mangnall]]></dc:creator>
            <pubDate>Thu, 10 Oct 2019 11:31:24 GMT</pubDate>
            <atom:updated>2019-10-10T11:31:24.330Z</atom:updated>
            <content:encoded><![CDATA[<h3>Diversifying teams with modern approaches to technical recruitment</h3><p><strong>Are once sought after standard academic routes into tech still relevant for modern companies?</strong></p><p>As a member of the People team at a growing software company I am involved in a lot of technical recruitment. From developers to testers and Azure experts, Spektrix is competing in a crowded market for top talent. Where we’ve found so much of our success is opening up recruitment and scrapping previous prejudices around education and training. In the rest of the article I will explore some of the alternative routes into technical roles, the advantages this brings to companies and teams and touch on other industries who are struggling to start exploring this.</p><p>There are new, exciting, flexible and accessible routes into technology careers appearing all the time. Simply put, you don’t need a Computer Science degree to become a successful developer, so why not pick a cheaper, quicker, more flexible option? Some of the key ones we’ve seen at Spektrix are intensive 12–16 week boot camps such as <a href="https://makers.tech/">Makers Academy</a>,<a href="https://flatironschool.com/"> Flatiron Schoo</a>l or <a href="https://northcoders.com/">Northern Coders </a>who teach students basic principles of Software Development and essential skills which can be transferred to the world of work. There are also a large number of organisations looking to up-skill certain individuals who are less likely to enter the industry via a traditional route. This includes <a href="https://www.codebar.io/">CodeBar</a>, a group of volunteers who organise free workshops around the UK &amp; United States to create a safe and collaborative environment for minority groups to learn together. Alternatively<a href="https://www.techreturners.com/"> Tech Returners</a> offer free part time courses designed for beginners and returners to help them reach a job ready level in just a matter of months.</p><p>What’s great is that the underlying values and principles around a lot of these organisations is making careers in tech accessible to those who may be excluded from joining in a traditional academic way. It is having a huge impact on the diversity of the talent within the industry. Furthermore these new candidates are demanding changes to outdated, homogeneous cultures and environments which in the past dominated the tech space. Naturally, workplace cultures have had to adapt to embrace this difference and become a safe, and inclusive place to succeed. It’s also quickly apparent that the skills these individuals bring from their previous careers make them stand out, and often excel quicker than those who have followed traditional educational routes and their different approaches to problem solving and creative thinking help teams to deliver success.</p><p>Other advantages which many of these ‘bootcamp’ courses build in from the beginning is employability. Part of the success of these organisations is partnering with employers in the areas they’re teaching. Building the schools’ credibility and networks in the community has been a key part of their growth. They are therefore aware of the tools needed to succeed within the roles at their partner companies and spend time preparing students for joining a role within tech, as well as providing key connections through job fairs or work sponsors.</p><p><a href="https://code.org/promote">Code.org</a> states that there are currently 493,270 open computing jobs nationwide, and yet only 42,969 computer science students graduated last year. The demand for technical talent means we need to modernise the way we educate people. The flexible, speedy and agile nature of learning mean alternative routes, outside traditional Computer Science degree courses, put engineers at an advantage simply given the demand from the industry itself.</p><p>From working in a position which influences recruitment, especially tech recruitment, I am excited as I see candidates’ backgrounds diversifying. Admittedly this talent joins us as junior engineers and when recruiting for senior positions it is much harder to attract or find the same levels of diversity. However, these essential changes to routes and education have made careers in technology more accessible and appealing for more people. Given time, the right mentors, and opportunities, these juniors will progress into our senior or leadership positions. Retaining and nurturing growing talent is incredibly important in an industry that moves as fast as tech, and critical for the success of any business. The future of tech looks brighter and more diverse because of this re-approach to learning and hiring. I urge other industries to look to modernise and diversify their hiring process and question whether old fashioned benchmarks such as; long academic degrees, Russell Group Universities, Private School education or regimented graduate recruitment is right for your company or the industry you work in as a whole.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b20ce8137c6b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/are-once-sought-after-standard-academic-routes-into-tech-still-relevant-for-modern-companies-b20ce8137c6b">Are once sought after standard academic routes into tech still relevant for modern companies?</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Five Tips to Kick-Start Product Discovery]]></title>
            <link>https://medium.com/spektrix-engineering/five-tips-to-kick-start-product-discovery-e24ae3cc0d83?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/e24ae3cc0d83</guid>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[product-development]]></category>
            <category><![CDATA[research]]></category>
            <category><![CDATA[product]]></category>
            <category><![CDATA[discovery]]></category>
            <dc:creator><![CDATA[Freddie Porter]]></dc:creator>
            <pubDate>Tue, 24 Sep 2019 11:00:46 GMT</pubDate>
            <atom:updated>2019-09-24T11:02:52.829Z</atom:updated>
            <content:encoded><![CDATA[<p><em>I’m Fred, a Product Owner at Spektrix. I’ve worked across various product opportunities, both technical and feature-based in scope. No matter how big or small the opportunity, discovery can be really overwhelming at the beginning. You’ve got lots of potential avenues to follow and undoubtedly many questions. To try and help you avoid this, I’m sharing some tips I’ve picked up along the way.</em></p><h3>1) Work out what you already know</h3><p>You’re having a hard time collecting your thoughts about what’s important to users or your business, and everything seems to be fighting for your attention at once. It’s time to focus and interrogate what you do and don’t know, and what questions are important enough to answer. Below is a method that has helped me get out of my <em>‘rabbit-in-the-headlights blues’</em>:</p><p>Start by listing all the information or questions you have in a simple table with three columns: <em>‘I know’, ‘I assume’ &amp; ‘I don’t know’</em>.</p><h4>I know</h4><p>This column should only include facts which you know to be true right now, and are able to back up with evidence such as research findings. If anything in the <em>‘I know’</em> column has even a whiff of assumption around it, then it should be exiled to the <em>‘I assume’</em> column until you’re sure you can’t disprove your assumption. Be as rigorous and as objective as possible here.</p><h4><strong>I Assume</strong></h4><p>This column should contain, as you might expect, notions which you assume or are assumed by others involved in your discovery. Perhaps there is a theory that you’ve got based on some quick observation of your website’s analytics, or some anecdotal feedback from a colleague about poor user experience. Maybe there is some evidence to support your hunch but you have questions about what that evidence means. Pop it all here.</p><h4><strong>I Don’t Know</strong></h4><p>This column should contain everything you don’t know, where everything is up for grabs and to be discovered. At the beginning of a project, it’ll likely be your largest column.</p><p>Even with all your knowledge split out into these three columns, things can still feel a little overwhelming as you likely don’t have a clear way to show you what to focus on first. At this stage, I find it a useful exercise to <strong>assign a risk factor to each of the items in the <em>‘I Assume’</em> and <em>‘I Don’t Know’</em></strong> columns. This risk factor is then a useful tool in helping me decide which of the items I should focus on researching further, so that I can either move them to the <em>‘I Know’</em> column, or disregard as unimportant for now.</p><ol><li><strong>Chance</strong>: What is the likelihood of my assumption or knowledge being wrong?</li><li><strong>Impact</strong>: What happens if my assumption or knowledge is wrong?</li></ol><p>You can assign a point system from 1–5 (1 being lowest chance or impact, 5 being highest chance or impact) to each item and then your risk factor is simply a calculation of both:</p><blockquote><em>Risk Factor = Chance x Impact</em></blockquote><p>Here’s a great example of how Trello and other ‘kanban’ style tool can be used to help you perform this exercise. Let’s imagine that we are starting an ice-cream cafe in a trendy part of London, and we want to ensure that we pitch a menu to appeal to our target market. We’ve done some research so far, but there’s lots more to be done. Early on in the project, one of our investors said that we should exclusively sell ice-cream in cones as most people love ice-cream in cones more than they like pots. We’re not so sure that this assumption is true so we put it into our <em>‘I assume’</em> column. In working out the risk factor, we assign these two values:</p><p><strong>Chance: 4/5 - </strong>We’ve not done any extensive research into ice-cream container preference.</p><p><strong>Impact: 5/5 - </strong>Cones are highly perishable (more so than sprinkles, for example) and therefore if we overstock without selling them we risk losing lots of money, and also risk putting off customers from purchasing our products if they can’t get the container they like.</p><p>In this case, <em>Risk Factor = 5x4 = 20</em>, which is very high, so we decide to test this assumption as quickly as possible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/842/0*1wdRkBdPAC0FdMPq" /></figure><h3>2) Move Smarter, not Faster.</h3><p>Discovery is all about getting answers as quickly as possible and working out what will help you move towards your discovery goal as soon as you can. Add to that the inevitable pressure you’ll be getting to deliver, it makes good sense to make the most out of your time at every opportunity. Be careful not to rush and compromise the quality of your work, but instead think about how you can maximise the value of your research in answering the most important questions you have.</p><p>When it comes to research and speed, Erika Hall puts it best in <em>Just Enough Research:</em></p><blockquote><em>Focus only on the essential user types</em></blockquote><blockquote><em>Deal with your data as soon as you get it</em></blockquote><blockquote><em>Involve your team in the analysis</em></blockquote><blockquote><em>Do the less important stuff later</em></blockquote><p>In other words, prioritise what is most important now (you can use your handy table you created above to help!) and test assumptions or theories using the data you get, as soon as you get it. Involve your team so that you’re sure to keep clear from your own bias when analysing, and kick less-pressing priorities down the road until they’re more important.</p><h3>3) Be resourceful.</h3><p>In a similar theme, you’ll never have the perfect set-up with the perfect conditions, or just enough time to get the most perfect set of data from the perfect participants. Life, business, and train delays all get in the way. Whilst you should absolutely treat your research scientifically and hold yourself (and others) accountable to do so, you’ll have to accept now that things won’t be perfect. Sorry!</p><p>So what can you do to be resourceful? Well, during a recent UX Crunch meetup, Sam O’Brien (User Experience Manager at RingCentral) floated the idea to apply weights to data collected from certain user types.</p><p>For example, imagine that to test our assumptions about cones vs pots for ice-cream, we’re taking on some market research. We are hoping to speak to our target markets, which we’ve split into <em>Primary</em>; <em>Secondary</em> and <em>Tertiary</em>. <em>Primary</em> is our most important market, <em>Secondary</em> our second-most, and <em>Tertiary</em> our third-most important.</p><p>Our Primary market is <em>‘Experiential Professionals’</em>. They’re people who like to try new things, live or work in the area close to our ice-cream parlour, and have disposable income to spend on boutique ice-cream. They’re most important to us as they are most likely to be our loyal customers and will spread word-of-mouth about our business.</p><p>Our Secondary market is <em>‘Weekend City-Breakers’</em>. They’re people who are on short holidays in London and want to check out the local areas to eat. They’re second-most important to us as they will help us diversify income streams, but are unlikely to return.</p><p>Our Tertiary market is <em>‘Social Media Super-Users’</em>. They’re people who are on the look-out for great content to share on their social media platforms, and therefore are likely to take pictures of our interesting flavour combos and share them online. They’re third-most important to us as they could potentially draw attention to our business online, but they are also unlikely to return.</p><p>Now that we’ve defined our markets and their importance to us, we can add a multiplier to data you collect against that user type to give that data additional weight in our research.</p><p>For our ice-cream parlour we have a list like this::</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/858/0*BTMqzEZm-f18dsY9" /></figure><p>In this example the markets <em>‘Experiential Professionals’</em> and <em>‘Weekend City Breakers’</em> are our most weighted, followed by <em>‘Social Media Super Users’</em>. We also include <em>‘Everyone Else’</em> with no weighting at all as, whilst we do want to take all opinions into account, we are trying to run a business and need to be realistic about who we aim our product at.</p><p>With this weighting in place, we can now begin to test our assumptions around cones vs pots safe in the knowledge that we can remain scientific with your approach and pay attention to our primary market most, whilst accepting that they may not end up dominating our research pool.</p><h3>4) Trust your gut… where appropriate</h3><p>It’s difficult not to feel like an imposter — especially when all your time is spent speaking to and learning about specialists. This is especially true in a B2B environment, where you’ll be dealing with users who know their own domain and therefore its needs well.</p><p>For example, imagine instead that our ice-cream parlour was a wholesale supplier instead. Our research would, in this instance, involve learning as much as possible about the businesses (parlours, restaurants, vans) we supply, so that we can be sure we’re providing a service that most suits their needs (and challenge those needs when it feels appropriate!)</p><p>As your research progresses, you’ll gain a nuanced understanding of the opportunity you’re discovering and of the user needs that you’re serving. Use this understanding and knowledge you have gained alongside your gut feeling to determine where to head next or what assumptions to disregard.</p><p>Beware though — it’s easy to conflate intuition with bias. Always try to ratify your proposed actions with users to ensure you’re on the right track.</p><h3>5) Document everything</h3><p>It goes without saying that you should keep a record of everything you do, find out, and track the decisions you make. Most importantly, you are not doing research for yourself. The expertise and knowledge you’re gaining during discovery are not only yours: they’re a business asset! Your documentation should be detailed enough so that someone else in your business could make decisions about how to proceed with an opportunity without your presence.</p><p>You should also aim to abstract your research documentation not only from you but also from any single opportunity. Where you are able to, ensure that your documentation has longevity. For example, with our ice-cream parlour market research, the information that we find out will be just as useful in guiding other opportunities further down the line as it was in aiding our discovery regarding cones vs pots. So, document in a way that allows this to happen. As a bonus point, having a clear log of the journey you’ve taken is also important for helping to tell a story about the opportunity, a key part of product management.</p><p>In summary, when starting out on your discovery journey remember to…</p><ul><li>assess what you know,</li><li>analyse data and make decisions quickly,</li><li>avoid perfectionism but keep it scientific,</li><li>trust your intuition</li><li>…and write everything down!</li></ul><p>Thanks for reading, let me know what you think and if you’ve got any more to add in the comments!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e24ae3cc0d83" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/five-tips-to-kick-start-product-discovery-e24ae3cc0d83">Five Tips to Kick-Start Product Discovery</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Spektrix’s first WomenHack night]]></title>
            <link>https://medium.com/spektrix-engineering/spektrixs-first-womenhack-night-bc2325d13101?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/bc2325d13101</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[women-in-tech]]></category>
            <category><![CDATA[diversity]]></category>
            <category><![CDATA[recruiting]]></category>
            <dc:creator><![CDATA[Emily Mangnall]]></dc:creator>
            <pubDate>Mon, 29 Apr 2019 15:38:16 GMT</pubDate>
            <atom:updated>2019-04-29T15:38:16.414Z</atom:updated>
            <content:encoded><![CDATA[<h3>Our first WomenHack night</h3><h4>A fun and engaging way to connect with women in Tech</h4><p>After almost doubling the size of our Engineering department in the last 12 months, we continue to grow our teams. It is now a company-wide commitment to focus on diversity, equality and inclusion. The diversity-focused recruiting night organised by WomenHack in London on the 4th of April was the perfect opportunity for us to tackle this goal.</p><p>Abbie <strong>— </strong>Software Engineer here at Spektrix<strong> — </strong>and I prepped our approach, packed as much Spektrix swag as we could carry, and headed off to the event.</p><h3>The Event</h3><p>The recruitment event, run by WomenHack and hosted by Vodafone, had 150 attendees. It kicked off with 5-minute keynotes from companies who sponsored the event (Giffgaff, Lloyds, Made.com, Red Badger &amp; Simply Business), followed by rapid interview sessions.</p><p>The Keynotes each followed a similar structure. Companies shared their experience trying to diversify their technical workforce and attract more females into technical roles. It was reassuring to know that companies of all sizes, industries and values are facing similar difficulties and have similar objectives to those we have at Spektrix.</p><p>The event became most enjoyable when the rapid interviews started. Our main intention was to chat with candidates with the experience and interest to be successful in our live roles. But we also wanted to raise our profile across the tech community, and reinforce our commitment to diversity and inclusion. So we did our best to turn our table Spektrix-green; filling it with branded notebooks, bags, pens, badges and of course job descriptions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*4zwE-MNJxTi5v0cD" /><figcaption>Who would refuse some Spektrix swag?</figcaption></figure><p>Attendance was very product, design and UX oriented which was perfect to fill our open Product Owner role. However, we would have liked to meet more back-end engineers, particularly senior developers.</p><h3>The Rapid Interviews</h3><p>The strict 5-minute speed dating style interviews had their pros and cons. They gave us just enough time to talk about our company and roles, find out about a candidate’s interest and gather their contact details. The time limit kept the conversations focused and to the point. It meant that conversation wasn’t too awkward when you inevitably were sat opposite someone who had no interest in your roles or company!</p><p>On the other hand, the time limit was difficult when a candidate was a potential perfect match. I can also imagine the frustration from the attendees’ perspectives. When you have done your research and identified the companies you’re interested in, you might want more time to chat with them in depth.</p><p>The floor did open up at the end of the night but this became a little overwhelming and not really useful either for hiring companies or candidates.</p><h3>Key Takeaways</h3><p>For a first attendance to the event, it was a huge success for Spektrix. As corporate-event rookies, we didn’t come ready with a fancy banner, but the conversations we had were really valuable. It was great to have the opportunity to introduce 100+ people to Spektrix and what we do.</p><p>As a B2B company that services theatres and playhouses, our brand is more established in the Arts sector than in the Tech world. So being able to tell our story to passionate women within the tech industry was great! The level of interest in Spektrix and our mission from candidates was fantastic. We definitely didn’t feel overshadowed by the large companies and household names of the brands surrounding us at the event.</p><p>The event allowed us to connect with several women from the event, giving us a chance to bring new people on board. Spektrix will definitely be seeking out similar opportunities in the future that can help us diversify and grow our teams!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*gTk_hCuANeYInlo8YR3kNQ.png" /></figure><p><strong><em>Spektrix is recruiting!</em></strong><em> Join our growing teams in London and Manchester and help build the SaaS platform pushing the Arts forward!</em></p><p><strong><em>Check out our </em></strong><a href="https://bit.ly/2IYnX9F"><strong><em>open positions</em></strong></a><strong><em>!</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc2325d13101" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/spektrixs-first-womenhack-night-bc2325d13101">Spektrix’s first WomenHack night</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[8 steps to make your CV sparkle]]></title>
            <link>https://medium.com/spektrix-engineering/8-steps-to-make-your-cv-sparkle-8fd85920d527?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/8fd85920d527</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[resume]]></category>
            <category><![CDATA[job-search]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[cv]]></category>
            <dc:creator><![CDATA[Dan Bass]]></dc:creator>
            <pubDate>Wed, 09 Jan 2019 14:01:01 GMT</pubDate>
            <atom:updated>2023-05-16T08:39:10.157Z</atom:updated>
            <content:encoded><![CDATA[<h3>8 Steps to Make Your CV Sparkle</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/464/1*LsiZ7s3HB1P0hHKZy9auJw.jpeg" /></figure><p>As we find ourselves in the technological age, it’s a candidate’s market, especially in the software engineering industry. Opportunities abound across diverse, exciting sectors. While saving time and effort on job applications is tempting, it’s crucial to remember the competition — your CV is your first impression for your potential dream job.</p><p>Having reviewed thousands of CVs and hired countless software engineers over the past decade, I hope you find these eight simple steps helpful in making your CV stand out.</p><p>Creating an impactful CV is challenging. It’s not part of our everyday routine. We need to channel creativity, articulation, and persuasive writing to catch the reviewer’s attention and present ourselves in the best light.</p><p>Conversely, it’s quite easy to construct an ineffective CV. A hurriedly assembled collection of generic statements, project descriptions, team interactions, processes, and a laundry list of every technology you’ve used can quickly become an indigestible waffle, especially for the reviewer with 20 more CVs to evaluate that day.</p><h3>1. Keep your CV concise and clear</h3><p>The best CVs convey their key points in a clear and succinct manner. Aim for a two-page limit.</p><p>Avoid cramming too much into these two pages. A simple structure that works includes:</p><ul><li>A brief introduction about you</li><li>Key skills</li><li>Recent roles first, highlighting five key achievements</li><li>For older roles, a line or two about significant achievements</li><li>A succinct overview of your education</li><li>Lastly, interests outside work… is there something about you worth sharing?</li></ul><h3>2. Include something memorable in your CV</h3><p>What will make your CV stand out?</p><p>There are CVs with clever touches that leave a lasting impression. Here are a few examples:</p><ul><li>Fascinating hobbies or facts about you “outside your career”</li><li>A compelling opening statement exhibiting a touch of humour</li><li>A splash of colour</li><li>The CV format does not need to be a PDF or DOC. How about a “.py” file for a Python developer or a “.cs” file for a C# developer position?</li></ul><h3>3. Tailor your CV for the role you’re applying for</h3><p>CVs that have been slightly customised for the role stand out.</p><p>No need to rewrite your CV for each role. From a solid foundation, create a few variations of your CV. Highlight the top 5 or 6 requirements and skills from the job specification, and make sure your CV addresses these points.</p><h3>4. Craft a compelling cover note</h3><p>A cover note is an excellent opportunity to show your enthusiasm for the company.</p><ul><li>What excites you about the company’s mission?</li><li>What inspires you about the role?</li><li>Why should you be considered above all other candidates?</li></ul><p>Avoid a generic cover note. Personalize it to show your understanding of the role and the company.</p><h3>5. Express your passion for your craft</h3><blockquote>“The WHY is the purpose, cause or belief that drives every one of us.” — Simon Sinek</blockquote><p>Nothing starts a CV off better than describing your “Why”.</p><p>For software engineers, our craft extends beyond writing code. We build something valuable and useful for customers.</p><ul><li>What are your principles for writing code that keeps your customers satisfied?</li><li>What practices do you adopt to ensure your code remains maintainable, efficient, and easy to deploy?</li></ul><h3>6. Highlight your key skills</h3><p>It’s tempting to list all the technologies you’ve used in your career but resist that urge.</p><p>Instead, list your top five skills — these could be technologies, programming languages, or techniques. More important than the programming languages is demonstrating continuous learning. What are you doing to grow and share your knowledge within the company and the broader community?</p><h3>7. Showcase your achievements</h3><p>Focus on the outcomes, not just tasks.</p><p>Consider the following questions:</p><ul><li>What measurable value did your projects deliver?</li><li>What challenges did you overcome in your work?</li><li>How did you improve non-functional requirements (like stability, scalability, and security)?</li><li>What feedback did you or your team receive?</li><li>What were your failures, and what lessons did you learn from them?</li></ul><h3>8. What are your hobbies and interests?</h3><p>One of my former managers told me, “To truly understand a person, look at the last few lines of their CV”. It’s fascinating to see who people truly are.</p><p>In conclusion, to make your CV shine…</p><ul><li>Keep it concise</li><li>Highlight your achievements</li><li>Emphasize your principles</li><li>Show a glimpse of your personal side</li></ul><p>Your efforts will pay off.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8fd85920d527" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/8-steps-to-make-your-cv-sparkle-8fd85920d527">8 steps to make your CV sparkle</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Delivering Tangible Value with Redis]]></title>
            <link>https://medium.com/spektrix-engineering/delivering-tangible-value-with-redis-f0ada1181b68?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/f0ada1181b68</guid>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[redis]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Freddie Porter]]></dc:creator>
            <pubDate>Tue, 31 Jul 2018 12:01:03 GMT</pubDate>
            <atom:updated>2018-08-01T13:56:26.670Z</atom:updated>
            <content:encoded><![CDATA[<h3>Delivering Tangible Value through technical projects</h3><p>I joined the Engineering Team at Spektrix as a Product Owner in March this year. It was an area of the business that I’d wanted to join for some time. Before, I’d been a Support &amp; Training Consultant for Spektrix. I answered phones, trained clients, delivered strategic projects and more. Before that I spent 6 years in the arts ranging from production to marketing. So, the idea that I would be a Product Owner in an engineering team was a new challenge, to say the least.</p><p>Upon joining I discovered the priority for my team was to “move session state from our SQL Server to another product”. Nothing quite like jumping in at the deep end. I’d seen engineers write the word ‘SQL’ on post-it notes, a fate to which I too was soon to be resigned. I had no idea what this word meant but my perspective would prove useful throughout the project.</p><h3><strong>What problem are we solving?</strong></h3><p>Since we’re talking about session state here, let me give you a little insight into how that works at Spektrix. When someone ‘hits’ Spektrix they create a session. This session state gets written into our SQL Server, in a database for session state. You can see the architecture for this below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/697/0*AES7QXXoArTyFQK-" /></figure><p>There are a few problems with this architecture:</p><ul><li>The Application and session database compete for resources on the same SQL Server instance.</li><li>There was a lot of I/O pressure on the SAN (Storage Area Network).</li><li>We often get SQL locks due to the way we hold seats off for customers whilst their session is still open.</li><li>The handling of session timeout in batches often failed or ran for a long time.</li></ul><h3><strong>What value are we creating?</strong></h3><p>We can now pinpoint the value that solving this problem will bring to the business. What’s important here is that the value must be tangible. That’s where my experience comes in; I know what the wider business wants to know. For example, it would be technically correct to say that solving this problem would:</p><ul><li>Reduce competition for resources on the SQL Server.</li><li>Reduce I/O on the SAN.</li><li>End SQL Locks due to session state.</li><li>Allow for better handling of session timeout.</li></ul><p>Yet, this isn’t particularly useful to know if you’re not an engineer. So how can we communicate value? We need to think of our audience and work out what is most relevant to them. We need more tangible value for our colleagues who are not in the engineering team. It’s more appropriate to say that solving this problem would:</p><ul><li>Ensure we can scale our business.</li><li>Ensure clients can scale their businesses.</li><li>Reduce the risk of slow-downs, especially during on-sales*.</li><li>Speed up sale of seats for in-demand events.</li></ul><p><em>*The term ‘on-sales’ here references the act of putting a new show, performance or event on sale to the public. These ‘on-sales’ result in spikes of traffic above general usage.</em></p><p>The value of solving this problem is quite easy to see. We can now propose this value to the business using a <a href="https://docs.google.com/presentation/d/11e5SVbUdLFbTsLgxYcdRf0RkJqlAiNs7QcVeCSXnsho/edit?usp=sharing">value proposition canvas</a>. This document is a way to show the problem we’ve found, the value in solving it now, and the risk in doing so. If the business thinks the value is worth delivering we prioritise the issue and a Dev team can begin to solve it.</p><h3><strong>Can you Dev it? Yes we can!</strong></h3><p>Once we — the development team — have the problem in our hands, we start on solving it. This starts with a kick-off meeting. We define the solution, create high level stories, key milestones, and outline desired outcomes and success metrics.</p><p>The solution which would deliver the value would be to host our session state in a location more suited to our needs than our current SQL database. We looked for other possible tools to use and settled on Redis. What is that, you ask? It’s an in-memory key value store and message broker — you can find out more information at their website <a href="https://redis.io/">here</a>. We chose Redis because it’s fast, it has a small footprint, operates only in-memory, and can send us messages. So, we started work on this solution.</p><h3><strong>When are we done?</strong></h3><p>To be able to measure our success and determine when we’ve delivered value, we create success metrics. Measuring our success throughout the project keeps us on track to deliver value. Our success metrics were:</p><ul><li>No problems caused by lock waits due to session state.</li><li>Reduction in writes on db due to session state.</li><li>Reduction in I/O on the SAN.</li></ul><p>Now we’ve got our framework for success in place, we can start work on delivering value.</p><h3><strong>How’s it going?</strong></h3><p>I’m asked this a lot. A key part of my role as Product Owner is communicating progress to the stakeholders. At Spektrix, we have many conduits for this such as:</p><ul><li>Weekly Stakeholder Updates.</li><li>Product Team Meetings.</li><li>Release Emails.</li><li>Sprint Digests.</li><li>Weekly Engineering Roundups.</li><li>Friday Demo &amp; Celebrations.</li></ul><p>For this project, the challenge was communicating our progress to the wider business. Given the niche problem on which we were working, I had to ensure that progress was tangible. In a Friday demo I introduced the problem, the value, and our proposed solution. I also needed to show our delivery plan, so stakeholders could benchmark our progress. So what was our delivery plan, you ask?!</p><h3><strong>Signed, Sealed, Delivered?</strong></h3><p>At kick-off, we broke down our solution into milestones. These milestones were points in the project where we would deliver tangible value. In the case of this project, this was in three stages:</p><ul><li>Simple session state management with Redis.</li><li>Improve our session timeout service.</li><li>Load-testing and cleanup.</li></ul><h4>Stage One:</h4><p>The first stage was to move the handling of our session state to Redis. We built a deployment pipeline to production with a skeleton service. This would handle expiry messages from Redis, and clean up any related data. We then enabled our main Spektrix solution to interface with Redis. We now had a working end-to-end simple session timeout service implementation using Redis in production. Enough value is now ready and rolled out to clients. We staggered client upgrades to the new way of handling sessions over two weeks. This enabled us to spot problems without putting the entire client base at risk. There were no problems with the simple session timeout service, so we moved forward.</p><h4>Stage Two:</h4><p>The next phase of work added resiliency to the simple session timeout service. We used Consul to handle service availability and deployed many timeout services via TeamCity and Chef. Our devops team created a Redis cluster, using Docker Swarm, of 6 nodes with 3 masters and 3 slaves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/881/0*teL8sL-csbwxHU8-" /></figure><p>If 1 node failed we continue to operate at 100% capacity in timing out sessions and cleaning up baskets. If 2 nodes failed, we could operate at 50%. The legacy service often failed or took longer than expected to complete. So, our improvements meant that we could restore confidence in our timing out of sessions. Now we had delivered this resiliency, we could move on to success measurement and cleanup.</p><h4>Step Three:</h4><p>The final step was a load-test was to ensure that the capacity of the system was not affected. We also wanted to remove the SQL server database we had stored session state in, as we would no longer need it. We tested system performance with usual system noise and, out-of-hours, pushed the system to failure. I’m pleased to say that our load-tests passed all previous records set!</p><h3><strong>Was it a Success?</strong></h3><p>Now that the project is over we need to ensure that it was a success and reflect upon where we could have improved. We must look back at our success metrics defined earlier in the process.</p><p><strong>They were:</strong></p><ol><li>No problems caused by lock waits due to session state.</li><li>Reduction in writes on db due to session state.</li><li>Reduction in I/O on the SAN</li></ol><p><strong>Have we achieved success?</strong></p><ol><li>There have been no problems caused by lock waits due to session state.</li><li>As we no longer have a session database , there are now no writes to this db due to session state.</li><li>We have now seen a 50% reduction in I/O on the SAN.</li></ol><p>Looking at these success metrics, it’s clear to say that this was without doubt a success. In communicating value, these success metrics aren’t clear outside of the development team. How do we communicate success here? The most important aspect is to think of your audience and what will be important for them. What’s important for our Sales team is different to what’s important to our Support team. For the sales team we can assure them that, due to this work, we can continue to scale our business and take on new clients. For our Support team, they’re pleased to know that they will no longer see seats in an unexpected state. For Account Managers, they care that our platform is more scalable and resilient. This means that our current clients can continue to grow their businesses too.</p><h3><strong>On Reflection</strong></h3><p>As part of every project at Spektrix, we include a ‘post-project’ retro to reflect on our work and how we can improve. This retro brings together anyone involved with the project across the business. We celebrated success and came away with actions for improvement. These were in response to the time it took to complete the project as we encountered issues releasing the timeout service.</p><p><strong>Actions included, but weren’t limited to:</strong></p><ul><li>Be more prepared to push back on assumptions or preconceptions.</li><li>Always be ready to ship.</li><li>Always challenge blockers that seem outside of your control.</li></ul><p>We’re confident that we will improve our ability to drive projects over the finish line and deliver value more continuously. This project was a huge success and will enable us to continue scaling.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*gTk_hCuANeYInlo8YR3kNQ.png" /></figure><p><strong><em>Spektrix is recruiting!</em></strong><em> Join our growing teams in London and Manchester and help build the SaaS platform pushing the Arts forward!</em></p><p><strong><em>Check out our </em></strong><a href="https://bit.ly/2IYnX9F"><strong><em>open positions</em></strong></a><strong><em>!</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f0ada1181b68" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/delivering-tangible-value-with-redis-f0ada1181b68">Delivering Tangible Value with Redis</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How a micro-optimisation in the .NET compiler broke our web application at runtime]]></title>
            <link>https://medium.com/spektrix-engineering/how-a-micro-optimisation-in-the-net-compiler-broke-our-web-application-at-runtime-7f418e5ded90?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/7f418e5ded90</guid>
            <category><![CDATA[dotnet]]></category>
            <category><![CDATA[testing]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[csharp]]></category>
            <dc:creator><![CDATA[Matt Thompson]]></dc:creator>
            <pubDate>Thu, 19 Jul 2018 13:01:01 GMT</pubDate>
            <atom:updated>2018-07-19T13:01:01.057Z</atom:updated>
            <content:encoded><![CDATA[<p>Recently we updated our builds to use Visual Studio Build Tools 2017, including MSBuild 15, instead of MSBuild 14. We expected nothing to go wrong with this. And getting to use the C#7 features, which we’ve had to ignore for awhile, was something nice to look forward to.</p><p>The commit went in, the build ran on the build server, unit and integration tests passed, and we’re ready for a release! During a last-minute bit of manual poking, we discovered a runtime error attempting to view any customer record when the customer had purchased a particular type of product.</p><blockquote><em>Uncaught Error:<br>Sys.Webforms.PageRequestManagerServerErrorException: Type ‘Spektrix.InterfaceModel.FixedSeries.MyFixedSubscriptions+&lt;&gt;c__DisplayClass6_0’ in Assembly ‘Spektrix.InterfaceModel’ is not marked as serializable.</em></blockquote><p>One thing stood out: the DisplayClass. Understanding why we were suddenly getting errors here relied on knowing what this thing is and why it’s here.</p><h3>Use of DisplayClass by the compiler</h3><p>When using a lambda that would reference some state and require a closure, the compiler generates a new class which holds the variables on the heap. This class contains the variables used in the lambda, and the lambda method itself. Generally, it’ll look something like this when decompiled:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/587/0*yArafZyFpaBrnA_O" /></figure><p>In any code using this, an instance of the DisplayClass is created and the state is assigned. Then an Action with a pointer to the method group on the instance is created, in place of the lambda.</p><p>However, when no state is required in a closure, the compiler would do something different. No DisplayClass would be generated, the lambda method would just be a method on the class in which it’s used:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/426/0*njcNsbyIe3hiZfDT" /></figure><p>But why DisplayClass? <a href="https://stackoverflow.com/a/14092456">This answer from Eric Lippert</a> explains that really, it’s just jargon:</p><blockquote>“The reason that a closure class is called “DisplayClass” is a bit unfortunate: this is jargon used by the debugger team to describe a class that has special behaviours when displayed in the debugger. Obviously we do not want to display “x” as a field of an impossibly-named class when you are debugging your code; rather, you want it to look like any other local variable. There is special gear in the debugger to handle doing so for this kind of display class. It probably should have been called “ClosureClass” instead, to make it easier to read disassembly.”</blockquote><h3>Why was this breaking for us now?</h3><p>For our back-office web application, we serialize all session state into Redis, including the current view state. This allows someone to leave the site and come back to exactly where they were, even if in the middle of a complex flow. Part of this serializable view state was an Action with no closure which was used to trigger data binding when something else somewhere on the UI happened. On upgrading to MSBuild 15 these Actions were no longer serializable.</p><p>Something had changed, and it was breaking a few unrelated areas of our product. All the breakages shared the same symptom — a lambda being serialized as part of the view model. Without finding out exactly why it was breaking, changing to a method group in the class fixed the issue. This meant that on MSBuild 15, the compiler was placing the lambda that didn’t require a closure inside a DisplayClass, while before it wasn’t.</p><p>Through some research we found that a micro-optimisation in the Roslyn compiler changed how delegates are invoked causing all lambdas to be part of DisplayClasses regardless of whether they relied on the state in a closure, as this was more performant. <a href="https://stackoverflow.com/a/30897727">This answer from Jon Skeet</a> explains exactly what we were seeing:</p><blockquote>“IIRC, it was found that invoking a delegate which was “backed” by an instance method was faster than invoking a delegate backed by a static method — which is the motivation behind the change.”</blockquote><h3>Finding and fixing</h3><p>We knew we couldn’t do what we were before. We wanted to be able to use the latest features available in C#, so we had to fix the places that were wrong. Doing some regex searches on the code to find any class marked Serializable that contained Action, Action&lt;&gt;, or Func&lt;&gt; properties found hundreds of classes. Narrowing down all places that would assign lambdas to them and then converting them to method group assignments got rid of the runtime exceptions.</p><h3>What did we learn from this?</h3><p>Legacy codebases can have a lot of skeletons in the closet. No one on the team would intrinsically know that a problem like this would occur from a version bump on MSBuild (and the related tooling). And if we had not caught it prior to release we would have had severe regressions in multiple places.</p><p>All levels of testing are important, and in this case a UI element callback being serialized into a 3rd party dependency was the failure. Naturally, in a codebase like ours, attaining 100% test coverage after the fact is prohibitive. So we’re focusing on fast, cheap rollbacks when unknown problems occur, as that is more valuable at this point in time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*B91gNjFj3djgrbcqLwXZJA.png" /></figure><p><strong><em>Spektrix is recruiting!</em></strong><em> Join our growing teams in London and Manchester and help build the SaaS platform pushing the Arts forward!</em></p><p><strong><em>Check out our </em></strong><a href="https://bit.ly/2IYnX9F"><strong><em>open positions</em></strong></a><strong><em>!</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7f418e5ded90" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/how-a-micro-optimisation-in-the-net-compiler-broke-our-web-application-at-runtime-7f418e5ded90">How a micro-optimisation in the .NET compiler broke our web application at runtime</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Want to spark change? Start with people!]]></title>
            <link>https://medium.com/spektrix-engineering/want-to-spark-change-start-with-people-c06b6bb7627a?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/c06b6bb7627a</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[project-management]]></category>
            <category><![CDATA[change]]></category>
            <category><![CDATA[work]]></category>
            <dc:creator><![CDATA[Ludivine Siau]]></dc:creator>
            <pubDate>Fri, 06 Jul 2018 13:34:59 GMT</pubDate>
            <atom:updated>2018-07-06T13:43:20.565Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*KeXzCxR_7APy02pzVgIW7g.jpeg" /><figcaption>Oh, do you?</figcaption></figure><p>Whatever your role is in your company, I bet you have already taken the initiative to change something around you. That’s because you care about your job and you want to leave things better than when you arrived.</p><p>Adding a new selection of snacks in the kitchen, organising a monthly team away-day, introducing a new development tool, improving a process… Big or small, change that matters never happens in a vacuum. It has an impact on people around you (that’s the point, right?)</p><p>These people might be your coworkers, sitting in different circles within your organisation: your team, your department, other teams, senior management, etc… Or they might be outside of your company: customers, users, partners, etc… You might be working with them on a day-to-day basis or catch up every few weeks. Or maybe you have never been in touch. They might not even know who you are! Whoever they are, the success of your initiatives will depend on how well you interact with them.</p><p>Taking the time to engage with people can feel like a burden when you want to progress fast. You might be tempted to skip that step and “just do it”. But if you want your work to make a positive difference, “doing it” won’t be enough. Working <strong>*with*</strong> people is what will take you over the line.</p><h3>Start by nailing down your objectives</h3><p>This is not the focus of this article. But before you can talk to anyone about your project, you need to be clear about what you’re trying to achieve.</p><p>What’s your intention? What’s the problem you’re trying to solve? What outcome are you aiming for? What’s your action plan? Be prepared to talk about these points. You won’t get anyone’s buy-in if you look like you don’t know what you’re doing.</p><h3>Find your stakeholders</h3><blockquote><em>A stakeholder is a person with an interest or concern in something, especially a business.</em></blockquote><p>The <em>something</em> is the change you want to make: cleaning up old git branches, adding a new column to your team’s JIRA board, improving your bugs prioritisation process, launching a new blog for your department, introducing OKRs as way to set goals for your company, etc… Regardless of the size of your project you will find at least one person who has an interest or concern in it.</p><p>The following questions might help you identify the people who have <em>an interest or concern</em> in your initiative:</p><ul><li>Whose job will you impact? Whose decisions rely on the area you want to change?</li><li>What perspective or information are you missing? Who can provide it?</li><li>Whose help do you need to make your project a success?</li></ul><h3>Find the right way to engage with your stakeholders</h3><p>There are a few things you need to consider in how you approach people in your project:</p><h4><strong>Pave the way for change</strong></h4><p>Each organisation, each team, each person reacts differently to upcoming change. Some might be very nimble and adapt to new information and context without a fuss. Others might need more time and help to prepare for change.</p><p>Take some time to think about how they might receive your initiative. Will it surprise them? Are they expecting the change?</p><p>Do everything you can to remove the element of surprise. Ease your stakeholders into change.</p><p>Taking a look at how they usually plan and implement projects might be helpful. It will give you clues on what they need to feel confident about change: information, lead time, ceremonies, checkpoints, approvals, documents… Make some compromise to follow the practices that will earn you their trust. It might feel like a detour or a slow down but it will actually remove a lot of roadblocks for your project. And build strong relationships for future collaborations!</p><p>And if you don’t know how your stakeholders prefer to approach change, ask them! It’s a good way to engage with them and show you want to do things the right way.</p><h4>Set expectations</h4><p>Your stakeholders will have a different perspective on the problem you want to solve, depending on their role, team, organisation… They might not even be aware of the problem! Start by bridging that gap: explain what you are trying to solve, ask for their take on it. You might learn a bit more about the problem and refine your plan.</p><p>Once you have a common understanding of the problem, discuss the outcomes you are targeting, how you plan to get there and what help you need from that person. Make sure you agree on who will be accountable and responsible for what.</p><h4>Listen!</h4><p>As much as you can, try to anticipate the questions and risks your stakeholders might raise. It does help to have some plan or elements of answers ready before you start a discussion.</p><p>But what matters the most is to be ready to actively listen. Invite people to give feedback. Show that you value their input and adapt your plan accordingly.</p><h4>Be respectful of prior contributions</h4><p>Keep in mind that someone built the thing you are trying to improve. Sometimes with love. At least with the best intention and care. Recognise the efforts people have invested in the area you are changing when your introduce your initiative to them.</p><p>Present change in a positive light. Don’t undermine or criticise what has been done in the past. Instead you can highlight that the context has evolved and raise the need for a different solution. Remove judgement from your representation of the problem and focus on how you want to make things better for the future.</p><h4>Adjust your timing</h4><p>Does your initiative align with your stakeholder’s current priorities? If not, you will have a hard time getting them involved. You will have better chances to get their help, feedback or approval if the problem you’re solving is as important to them as it is to you.</p><p>Take your stakeholders’ priorities into account and adjust your timing to avoid struggling with their availability. If you push your initiative forward now, how will it impact their job? Will you disrupt them? Will they have time to feedback or will they feel rushed?</p><p>Consider slowing down or delaying to give them the time to get on board. And in case you have to progress without them, think about how they might feel about that and how it will affect your project down the line.</p><p>Using the right channel to introduce your initiative is also crucial. Choose the one that fits your audience’s workflow and allows them to give your message the attention it needs.</p><p>All the above may look very heavy when you’re trying to put in place a small change, but it doesn’t have to be. It’s not about asking permission for everything you want to do. Sometimes, it can be as simple as having a quick chat with the right person.</p><p>Whatever you’re trying to achieve, make sure you pause to think about who you need to get on board and how. Based on your stakeholders’ level of interest and concern, you can tune how involved your approach is.</p><p>To recap, here’s my personal cheatsheet to get people on-board with my projects:</p><ul><li>No change without a clear goal!</li><li>Seek feedback, from the start, and throughout</li><li>Don’t rush (people) into change</li><li>Be transparent about your intent, your plan and your progress</li><li>Put a positive spin on change</li><li>Show empathy</li></ul><p><strong>What would you add to that list?</strong> Share your own experience working through change with people. I’d love to hear from you!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*B91gNjFj3djgrbcqLwXZJA.png" /></figure><p><strong><em>Spektrix is recruiting!</em></strong><em> Join our growing teams in London and Manchester and help build the SaaS platform pushing the Arts forward!</em></p><p><strong><em>Check out our </em></strong><a href="https://bit.ly/2IYnX9F"><strong><em>open positions</em></strong></a><strong><em>!</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c06b6bb7627a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/want-to-spark-change-start-with-people-c06b6bb7627a">Want to spark change? Start with people!</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[DeltaV Conference: 3 speakers who changed our perspective on Web Performance]]></title>
            <link>https://medium.com/spektrix-engineering/deltav-conference-3-speakers-who-changed-our-perspective-on-web-performance-adc1ecc3c7df?source=rss----f9f911ef2ee---4</link>
            <guid isPermaLink="false">https://medium.com/p/adc1ecc3c7df</guid>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[web-performance]]></category>
            <category><![CDATA[front-end-development]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[conference]]></category>
            <dc:creator><![CDATA[Abigail Jackson]]></dc:creator>
            <pubDate>Tue, 12 Jun 2018 11:23:08 GMT</pubDate>
            <atom:updated>2018-06-12T11:25:52.871Z</atom:updated>
            <content:encoded><![CDATA[<p><em>DeltaV is a new conference focused on web performance in all its forms. Perry Dyball (Spektrix/London Web perf), Jo Franchetti (Samsung Internet) and Simon Hearne (Akamai) organised the first edition in conjunction with White October events and a few other sponsors. The event took place over 2 days in central London, with a strong focus on diversity and inclusion, both in terms of delegates and running themes. Underrepresented groups in tech, and those facing economic or social hardship, were offered easy access to the conference through a scholarship scheme.</em></p><p><em>We had the chance to attend as guides, helping the audience and speakers feel welcome and get the best out of the conference. We got a lot out of the two days and we’re excited to bring back what we learned to our teams and Spektrix.</em></p><p>The topic of web performance is relevant to everyone across any business. It was quite an eye opener to see that it covered more than ‘just’ front end optimisation.</p><p>Yes, there were talks on topics like trimming down your use of third parties. But web performance isn’t relevant only for engineers. Instead of only focusing on techy bits, the event brought attention to user interaction and engagement of users with the platform.</p><p>The success of a product depends on how well the business embraces the constant need to optimise the user experience, whether it’s about accessibility or platform performance. That was the overarching theme that all the speakers were trying to get across.</p><p>One theme particularly resonated with us: the impact that diversity and inclusion within a tech company or engineering team have on performance. It appears in the three talks we have chosen to explore below.</p><h3>Hunting the Unicorn</h3><p><em>Related by Abbie Jackson, Software Engineer</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*ZlzlWxegVBjel5LzbcXDuA.jpeg" /><figcaption>Tammy Everts — Photo by Garrett Coakley for White October Events (<a href="https://www.flickr.com/photos/whiteoctevents/sets/72157696806850385">DeltaV Conference 2018 on Flickr</a>)</figcaption></figure><p>Tammy Everts of SpeedCurve opened day 2 with her talk “Hunting the Unicorn”. She discussed why it is important to use web performance metrics relevant to the business or end users.</p><p>Tammy talked about how we need to consider “web stress”, which she described as being pretty much anything that affects accessibility to your website or app. This can relate to many things: accessibility issues because of a disability, bad UX, or even how a drunk person struggles to navigate your application!</p><p>Without regurgitating the whole of Tammy’s talk (you should read <a href="https://speedcurve.com/blog/web-performance-monitoring-hero-times/">Tammy’s own blog post</a>), the gist of it was that the metrics you choose need careful consideration and are dependent on your business context. There won’t ever be a “perfect” metric, but engineers should try to use metrics that identify the cause of web stress.</p><p>How can any of this apply to Spektrix? Perhaps a useful metric for Spektrix to measure could be, “how quickly can a client set up an event via the admin interface?”, or, “how long does it take a customer to buy a ticket?”.</p><p>Whatever metric we use, customisation and relevance to the end users is key in your search for hunting the unicorn.</p><h3>Debugger; for developers</h3><p><em>Related by Alessandro Signoretti, Software Engineer</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*OWoFjg5QaccflHf14DQmjA.jpeg" /><figcaption>Denys Mishunov — Photo by Garrett Coakley for White October Events (<a href="https://www.flickr.com/photos/whiteoctevents/sets/72157696806850385">DeltaV Conference 2018 on Flickr</a>)</figcaption></figure><p>In his emotional theatre monologue, Denys Mishunov guided us on a journey to understand how common “bugs” in us as human beings and engineers can affect performance.</p><p><strong>Engineers tend to be Perfectionists.</strong> We keep repeating the mantra “my product should be perfect!”, which is the way we have to ensure we can be proud of our work. The healthy way to achieve this goal is to try to frequently release something that is ‘good enough’ and then move step by step towards “perfect” code. On the other hand, unhealthy perfectionism can lead us to procrastination. Paying too much attention to small, unimportant task makes us waste time. It lowers our confidence to release and increases complexity and bugs.</p><p><strong>The Impostor syndrome</strong> is a bug that affects most engineers. It manifests when you think your success is not due to your talent and hard work, or when you feel less competent than other people think you are.</p><p>Then we feel the need to catch up and consume a lot of information (articles, videos, blog posts, twitter etc). This can lead to us feeling overwhelmed and hence frustrated as we can’t keep up.</p><p>How do we fight impostorism? By embracing it! It’s normal to feel that way when you are learning new things. Compete with yourself, track your own progress, rather than competing with others. Understand your own achievements.</p><p><strong>Long Hours </strong>are another enemy of developers. When we are highly engaged in achieving our goals, trying to meet an important deadline or complete a challenging project, we put more time into work. This “Hard Working” approach can improve our productivity only when we keep it going for a short period of time.</p><p>However, when working long hours becomes the norm and work takes over our days and thoughts, it only takes a few weeks to see productivity decreasing.</p><p>This unhealthy behaviour feeds a vicious circle. Our brain gets tired, we become more prone to errors, which slows our progress down and makes us feel like we haven’t accomplished enough. So what do we do? We put more hours into work.</p><p>The stress generated by this routine can slowly lead to mental health problems and even put our lives at risk (deaths related to overworking have actually become a social problem in Japan and China).</p><p>As Harvard University suggests to their students, we need breathing room from our human bugs. We need more happy and healthy developers, so let’s acknowledge our problems and fix them right away: debug yourself often and push your changes into production!</p><p>Taking inspiration from this at Spektrix, we should keep pushing more for pair programming as it can help to get things done until the “good enough” point, without going overboard. It also gives us a platform for honest feedback to drive personal development, and a safe space to share fears and doubts about our performance. And let’s keep taking care of ourselves and discourage long hours.</p><h3>Speeding up without slowing down</h3><p><em>Related by Tom Loake, Front End Engineer</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*jey5n5r8SBJSH1sGrL13Kg.jpeg" /><figcaption>Rhys Evans — Photo by Garrett Coakley for White October Events (<a href="https://www.flickr.com/photos/whiteoctevents/sets/72157696806850385">DeltaV Conference 2018 on Flickr</a>)</figcaption></figure><p>Rhys Evans gave a fascinating overview of the processes behind ft.com.</p><p>The talk started from the point where their online application had become a bloated monolith, making it extremely hard to release. People were using hacks, tag managers, etc. to avoid having to wait for their one month big bang release cycle.</p><p>New tech teams were formed and they insisted on total control of what ended up on their site. They performed a major re-architecture and moved to a microservices-powered back end, a componentised front end and a significantly faster release cycle. These are all great things but each caused their own issues. More latency due to more microservices, a need for more monitoring, and asset churning leading to difficulties in front end caching, to name just a few.</p><p>What was interesting to me is that at ft.com, they felt like they went too far. The tooling they put around their dev services as well as processes they put in place made it harder to write code. It also seemed like they lost perspective on where the simplest wins were and had to work to get back to achieving them.</p><p>Hearing them talk about how to solve these issues was super interesting. It was great to hear a talk where people were open about what went wrong with their solutions to problems.</p><p>DeltaV was an eye opening experience and a great introduction to the web performance community.</p><p>It has already helped us frame our conversations from a performance perspective. It’s given us the tools to think about and communicate why that matters in Spektrix.</p><p>The conference was friendly, it was easy to talk to people, and the food was brilliant! We met some interesting people and hopefully got the Spektrix name out there a little.</p><p>It was also great to see that money was raised for Code Club, Code Bar and Code Your Future, all three inspiring initiatives that were represented at DeltaV.</p><p>We’re all looking forward to next year’s edition!</p><p><em>Abbie, Alessandro &amp; Tom</em></p><p><strong><em>Spektrix is recruiting!</em></strong><em> Join our growing teams in London and Manchester and help build the SaaS platform pushing the Arts forward!</em></p><p><strong><em>Check out our </em></strong><a href="https://bit.ly/2IYnX9F"><strong><em>open positions</em></strong></a><strong><em>!</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=adc1ecc3c7df" width="1" height="1" alt=""><hr><p><a href="https://medium.com/spektrix-engineering/deltav-conference-3-speakers-who-changed-our-perspective-on-web-performance-adc1ecc3c7df">DeltaV Conference: 3 speakers who changed our perspective on Web Performance</a> was originally published in <a href="https://medium.com/spektrix-engineering">Spektrix Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>