<?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 Ronald Ashri on Medium]]></title>
        <description><![CDATA[Stories by Ronald Ashri on Medium]]></description>
        <link>https://medium.com/@ronald_istos?source=rss-5ca576794be4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*CWYomOJq13HZ-rR1oTysFA.png</url>
            <title>Stories by Ronald Ashri on Medium</title>
            <link>https://medium.com/@ronald_istos?source=rss-5ca576794be4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 12:39:09 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ronald_istos/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[Chatbots are dead, long live conversational apps]]></title>
            <link>https://medium.com/@ronald_istos/chatbots-are-dead-long-live-conversational-apps-16500ff480dd?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/16500ff480dd</guid>
            <category><![CDATA[conversational-ai]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[conversational-apps]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Thu, 16 Jan 2020 09:13:04 GMT</pubDate>
            <atom:updated>2020-01-16T12:19:20.090Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dE1Ye-3G_EGZkAd8" /><figcaption>Photo by <a href="https://unsplash.com/@kellysikkema?utm_source=medium&amp;utm_medium=referral">Kelly Sikkema</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Words matter. They frame our thinking and influence our decisions. In the conversational space, we have an abundance of terms and it’s time we start challenging how they affect the field and our ability to effectively communicate benefits to others.</p><p>In particular, I think the term chatbot has outlived its usefulness. I had already <a href="https://medium.com/@ronald_istos/chatbots-have-an-identity-problem-its-time-we-got-things-straight-fd0d3ac3fbb1">written about this in 2018</a>, but that was with the hope that we could get clarity on the term chatbot itself. Instead, since then, things have gotten more complex and there is a proliferation of other equally vague terms that have joined the party. Terms such as virtual assistant, intelligent assistant, AI chatbot and conversational agent are used both interchangeably or as indications of different approaches, making it all difficult to untangle. It is time we explicitly worked to change how we describe what we do in the conversational space.</p><h3>What we know we know</h3><p>With a few years of commercial exploration behind us we should be in a position to accept a few ground truths. Namely:</p><ol><li>Artificial intelligence techniques may or may not play a role in the development of a solution. There is no “true” chatbot that uses only natural language processing or only rule-based matching. There are a number of issues that come into play when designing a conversational product and the only true measure is the final utility for users.</li><li>Solutions can and should adapt to different mediums/platforms. There is no definitive proof that we should only design for natural language or only use buttons or other form elements. It should be a choice based on what would best serve a user in a specific context.</li><li>There is nothing intrinsically positive about a solution that is able to chat about random subjects or fool the user into thinking they are chatting to a human. Similarly, we do not know enough about which personality or behavioural traits (or simulation of behavioural traits) would work best in a specific context. There is interesting work taking place and we are learning more every day and signs point to a nuanced set of parameters.</li><li>Saying that a virtual assistant is different or better than a chatbot is a completely arbitrary choice depending purely on a very context-specific definition of those terms designed to support the desired marketing outcome. <a href="https://www.forbes.com/sites/cognitiveworld/2018/12/23/yes-chatbots-and-virtual-assistants-are-different/#9394c086d7d2">These</a> sort of articles don’t help.</li></ol><p>Now, how do we get out of these arbitrary definition wars and purity contests and focus on what is really important?</p><p>First off, and most importantly, the problem we are trying to solve is not how to make computers able to converse like humans. That is an exciting research issue but not a goal for a single product. When building a conversational product there is a wide range of issues to tackle and having conversations that are as human-like as possible is an option to be driven by a clear understanding of what benefits it brings. There is much work we still need to do to better comprehend what type of personality or behaviour in a conversation will work best in what situation.</p><p>Second, the things that we build need to do much more than just talk (or chat). They are supposed to help people solve problems faster and more enjoyably than other options (otherwise we should be using a different option!). Conversational solutions need to integrate with other services, adapt to changing interfaces and platforms and always make sure they are effectively helping the user complete their goal.</p><h3>Conversational applications</h3><p>In short, what we are trying to build are applications. Just as we always have. Software that helps humans complete a task. What is different from all the other applications out there is that these apps are predominantly conversational.</p><p>So let us just call them that. Conversational applications.</p><p>It is primarily an interaction-style choice. One sufficiently new and interesting enough to require a specific sub-category in the application space. A choice that should be clearly justified because of the value it provides to the user.</p><p>We choose to build a conversational application because we have concluded that a conversation is the most efficient way to resolve the problem and/or a conversation is the most user-friendly way to interact through a specific platform.</p><p>We recognize that conversational applications have a specific and separate set of challenges they need to solve when compared to other types of applications. They range from AI challenges (e.g. how to interpret natural language) to UX (e.g. what conversational patterns make sense) to software engineering (e.g. how to test conversational applications). The only measure of whether an approach is right or wrong is whether users are happier. Not whether we are using deep learning, or are voice-first, or are buttons only.</p><p>As a community of practitioners building these things, it is useful to define what we are doing clearly. I think the most inclusive thing we can say is that we are building conversational applications. We can then focus on describing the rich and varied ways in which conversations can develop from voice to text to gestures and anything else we can think of. Equally, we need to start identifying what types of conversations work in what contexts, rather than assume that more personality or more engagement is always better.</p><p>Let’s give the term chatbot a break and return it to its rightful owners. Long live conversational applications.</p><p>At <a href="https://greenshootlabs.com/">GreenShoot Labs</a> we are building an open-source framework for creating conversational applications called <a href="https://opendialog.ai/">OpenDialog.ai</a>. I</p><p>f you are interested in how AI, Data and Messaging Platforms will influence how we do work check out my book “<a href="https://bit.ly/aiworkplace">The AI-Powered Workplace</a>” (Apress, 2020)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=16500ff480dd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building AI software: The working programmer’s conundrum]]></title>
            <link>https://medium.com/hackernoon/building-ai-software-the-working-programmers-conundrum-71dbd1bee93d?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/71dbd1bee93d</guid>
            <category><![CDATA[ai-software]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[programmer-conundrum]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Thu, 30 Aug 2018 17:25:34 GMT</pubDate>
            <atom:updated>2018-08-30T17:56:13.091Z</atom:updated>
            <content:encoded><![CDATA[<h4>What on earth does it mean to “build AI software”?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xWGWcVOV8ioH9io6" /><figcaption>This company is going to go AI-first on all its tech and that will cut costs, increase revenue and the world will be amazing.</figcaption></figure><p>You’ve been working on your company’s software for years. You know all the ins and outs, the paths that lead nowhere, the tripwires to hop over and the avalanche-prone areas where you must tread lightly.</p><p>Despite everything, this delicate balancing act of a hodgepodge of software (half of which you would have never picked yourself but “strategic alliances” dictated otherwise) runs all the internal infrastructure and external consumer-facing tools just fine. You sleep easy (most nights) and you have a plan on how to get to a better state piece by piece, sprint after painful sprint. It will be tough, but you will get there because you actually care about your craft.</p><p>Yesterday, though, the management team called you in. This company is waking up, they say. It’s getting with the times. It going to innovate itself out of the downward spiral it’s in. They have just two words for you. <em>Artificial Intelligence</em>. This company is going to go AI-first on all its tech and that will cut costs, increase revenue and the world will be amazing. They’ve been told this is what happens and they have bought the dream. All they need you to do is come up with a fail-fast, iterate constantly, scale instantly plan that will get everyone there. Report next week.</p><p>You go back to your desk and think: “What exactly is AI anyway? What on earth does it mean to “put AI in software” or to “build AI software”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rsvbNTC6wRN7ES1E" /></figure><h3>“What on earth does it mean to build AI software?”</h3><p>This, I think, is the working programmer’s conundrum (with regards to AI at least!). It’s not that we are concerned about change as programmers. We thrive on change. We are inundated daily with new ideas, new concepts, new frameworks and we have become very adept at understanding how to deal with them, what to ignore and what to take into consideration. AI, however, takes this problem to a whole new level.</p><p>It doesn’t help that whenever you start reading on the subject people start mentioning the need for PhDs and advanced math.</p><h3>I Am Devloper on Twitter</h3><p>You say: &quot;We added AI to our product&quot; I hear: &quot;We added a bunch more IF statements to our codebase</p><p>Worst still are those who say that AI is nothing to be too concerned about. It’s just business as usual. It’s a better recommendation algorithm, or some bit of fancy data analysis thrown in with a pretty visualisation perhaps. These are the same people that say they downloaded <a href="https://www.tensorflow.org/">Tensorflow</a> and built something cool in just a day.</p><blockquote>The biggest trick AI managed to play on working programmers is to make them think that it doesn’t exist.</blockquote><p><strong>We need to start thinking of AI as a fundamentally different way of building software.</strong> The ability of the software to <strong>learn</strong>, <strong>adapt</strong> and <strong>autonomously make decisions</strong> is placed front and centre with AI. That needs to be the starting point.</p><p><strong>We need a software engineering design methodology that places those concepts and those abstractions at the centre</strong>. From there we build out. From there we can start evolving the design until we get to the point were we start falling back to the tools we already have like object-orientation, or domain design or whatever is your preferred methodology to express a software design idea.</p><p>I’ve started writing about it <a href="https://hackernoon.com/building-ai-software-data-driven-vs-model-driven-ai-and-why-we-need-an-ai-specific-software-640f74aaf78f">here</a> in the context of data-driven vs model-driven AI and I will be expanding on some possible routes in the future. In the meantime it would be great to hear other thoughts. Do we need a different set of abstractions? Are the tools currently available to programmers enough to design and think about AI software?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=71dbd1bee93d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/building-ai-software-the-working-programmers-conundrum-71dbd1bee93d">Building AI software: The working programmer’s conundrum</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building AI software: Data-driven vs model-driven AI and why we need an AI-specific software…]]></title>
            <link>https://medium.com/hackernoon/building-ai-software-data-driven-vs-model-driven-ai-and-why-we-need-an-ai-specific-software-640f74aaf78f?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/640f74aaf78f</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ai-development-paradigm]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Fri, 24 Aug 2018 16:46:11 GMT</pubDate>
            <atom:updated>2018-08-24T16:46:11.634Z</atom:updated>
            <content:encoded><![CDATA[<h3>Building AI software: Data-driven vs model-driven AI and why we need an AI-specific software development paradigm</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ShvFJOah9_pncIXM" /></figure><h3>Introduction</h3><p>Current, industry-led, interest in artificial intelligence is almost entirely focussed on <em>data-driven</em> AI. The reasons are easy to understand. Cheap data storage, fast processors and advancements in neural net algorithms and other data-centric techniques have made it possible to extract huge value out of data. We can build systems that can predict what will happen next based on what they’ve seen so far, very efficiently. Their performance is at times even better than that of a human being. The focus on data-driven AI is such that people have gone as far as labelling data-driven machine learning “<a href="https://www.oreilly.com/ideas/what-machine-learning-means-for-software-development">the part of AI that works</a>”</p><p>How true is that statement though? Is there really nothing that other parts of AI can offer us? Is there nothing coming out of all those decades of research in logics, planning, semantics or expert systems that can help us build a better autonomous system? As you probably have guessed, I think the answer is a resounding “no”. Furthermore, I think that <strong>unless</strong> we effectively combine data-driven AI with model-driven AI we cannot build successful AI systems.</p><p>Without appreciating what a more structured AI model brings to the table we risk re-inventing the wheel over and over again and missing out on very useful tooling. In order to effectively combine different AI tools though, we need ways of understanding and describing the architectures of autonomous systems — <em>we need a software engineering paradigm specifically focussed on AI software</em>.</p><p>Before we dive into the whys of of all of this let’s quickly make sure we are working of the same definitions of data-driven and model-driven AI.</p><h3>Model-driven and Data-driven AI</h3><p>For the purposes of this discussion, artificial intelligence refers to the tools, techniques and methodologies we use to automate processes. If you want a more nuanced understanding of what is artificial intelligence I’ve explained <a href="https://hackernoon.com/making-artificial-intelligence-work-for-you-part-1-what-is-ai-dd7512058e0e">here</a> how I use agent-based engineering concepts to derive a definition that is helpful for <em>building</em> systems.</p><p>In the context of process automation, we have, broadly, two ways of building systems that can understand what they are supposed to do next.</p><h4>Data-driven AI</h4><p>The data-driven way focusses on building a system that can identify what is the right answer based on having “seen” a large number of examples of question / answer pairs and “training” it to get to the right answer.</p><p>There are lots of different ways of doing this, with perhaps the most popular being using neural network algorithms in their various forms.</p><p>The necessary ingredients for this approach are an appropriately large dataset that, crucially, is also correctly labelled. No small feat. However, if you do get to the point where you have enough, correctly labelled, pictures of cats you can train software by “showing” it those images and letting it know when it “guessed” it correctly or not. After many (millions) of training cycles it will “learn” to get it increasingly right.</p><p>The strength of this approach is that it does not depend on a human accurately describing through a set of rules when something is a cat (e.g. it’s cute, it has a round face, it’s furry, it’s very popular in internet memes, etc). We now know, having tried for decades, that we are terrible at actually capturing in a set of explicit rules these sort of things. With data-driven AI the system learns “on its own” when something is a cat based on the training data we gave it. The more and more varied the training data the better our system can be.</p><h4>Model-driven AI</h4><p>Model-driven AI (or symbolic AI), instead, attempts to capture knowledge and derive decisions through explicit representation and rules. In a model-driven world, a cat would be explicitly represented as a four-legged animal, with two eyes, a nose and a mouth that is furry (except when not) and that is relatively small (except when not), etc. A model-based system would look at an image, deconstruct it into lines and shapes and colours and the compare against the set of rules we’ve supplied about how lines and shapes and colours combine in the world to give us different animals.</p><p>You can immediately see why this is not a very good way of building a system to recognise a cat. There are so many different rules and exceptions to those rules that we can’t capture all of them. More fundamentally, perhaps, we as humans don’t actually know how we do it. How can we build a system that does it explicitly if we can’t even describe what we do when we decide something is a cat.</p><p>These types of examples are why model-driven AI can easily get dismissed. It is not a good fit for many different situations. The question, however, is whether there are situations that <em>are</em> a good fit for explicit models and whether we can have systems that are purely model or data-driven or whether we gain more from combining the two. So let us explore that next.</p><h3>When is it useful to combine data and model-driven AI</h3><p>Model-driven AI represents the attempt to capture our understanding of how the world works through explicit representation and rules. If the bit of the world we are trying to capture is “how you can identify a cat in a picture”, it is not a good fit. However, there are many domains that are highly codified, explicitly defined and can be captured in a model. Margaretta Colangelo puts it elegantly <a href="https://www.linkedin.com/pulse/why-small-data-important-advancing-ai-margaretta-colangelo/">here</a> where she discusses the value of small data as a complement to big data:</p><blockquote>“Everything was small data before we had big data. The scientific discoveries of the 19th and 20th centuries were all made using small data. Darwin used small data. Physicists made all calculations by hand, thus exclusively using small data. And yet, they discovered the most beautiful and most fundamental laws of nature. Moreover, they compressed them into simple rules in the form of elegant equations. Einstein championed this with E=mc² . Although it’s estimated that perhaps 60% to 65% of the 100 biggest innovations of our time are really based on small data, current AI developments seem to focus mostly on big data, forgetting the value of observing small samples.”</blockquote><p>If a model can be derived it can provide the most efficient path from question to answer.</p><p>Even when the model is not completely perfect it can still get us closer to an answer. Consider, for example, the problem of trying to build a conversational agent that is trying to help a user solve a problem in a specific domain — say the domain is “how to train your cat”.</p><p>Imagine that on the one hand, we have an NLP system that allows us to deconstruct what the user said, for which it often combines data-driven AI with explicit grammatical models of a language (model-driven AI). On the other hand, we have information collected from cat trainers about how one should go about training their cat in a number of different situations and for different types of cats. Let’s call the NLP system the generic layer and the knowledge we collected the domain layer. The generic layer is going to provide the “smarts” that our conversational agent uses to interpret what the user says and reduce it to a specific request for action (or intent). For example, the user might say something such as “I would love to to learn how to teach my cat to do tricks”. The NLP system can map that to a user intent such as “train cat to perform tricks”. With that intent available we now need to extract useful information from our domain layer around what tricks would be the most suitable for the owner in question and their cat. We need a model of that specific owner, their cat and then a set of rules that will allow us to come up with an appropriate training plan. How would we go about building such a model of our world?</p><p>While there is a lot of focus spent on the generic layer many existing systems follow relatively unstructured or “home-made” ways of dealing with domain models and reasoning over the domain layer. Instead, it is important that software engineers start thinking of what a complete AI system looks like and what are the other tools coming from AI research and related fields that can help. This is exactly where model-driven AI helps. For this example work from <a href="https://en.wikipedia.org/wiki/Ontology_(information_science)">ontologies</a>, <a href="https://en.wikipedia.org/wiki/Inference">inference</a> and <a href="https://en.wikipedia.org/wiki/Automated_planning_and_scheduling">planning</a> can all help us build a better system.</p><h3>A new (old) paradigm — agent-based software development</h3><p>In a more general sense, we need to think about the entire system as made up from a variety of sub-systems. Some sub-systems will be data-centric while others will depend on explicit models. While current focus on the data-driven layer is understandable, a wider focus on the entire system is necessary. However, part of the challenge of combining such sub-systems and avoiding re-inventing the wheel its that we have no widely used software development paradigm that deals explicitly with autonomous systems.</p><p>We should be able to reason over the entire architecture we are using and be able to refer to a library of architectures and patterns that allow us to solve common problems efficiently. Just like object-oriented programming helps with dividing a problem into smaller pieces and provides us with patterns to deal with common issues, we need to better comprehend what it means to build intelligent systems and what are the common patterns we can refer to.</p><p>The good news is that while there is no <em>widely used</em> paradigm, that does not mean that such a paradigm does not exist. From within artificial intelligence research, <a href="https://en.wikipedia.org/wiki/Intelligent_agent">agent-based software development</a> can help us reason about the system as a whole and about the interactions between the various subsystems.</p><p>Agent-based software development deals with exactly these issues. It offers a variety of different architectures, both for individual agents and for systems with multiple agents interacting. One of my personal objectives moving forward is to talk about these different approaches, building on existing work, and share tools that can help us easily translate agent architectures into working software.</p><h3>Towards a more mature AI engineering ecosystem</h3><p>Artificial intelligence tools are here to stay. As the technology matures we need to start thinking about how our understanding of the software engineering aspects of it need to mature to allow us to capture and share knowledge and experience. There is a lot of existing work within artificial intelligence research that can help with this. The end result is that we better understand how intelligent systems can be built, using both data-driven and model-driven AI to solve problems.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=640f74aaf78f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/building-ai-software-data-driven-vs-model-driven-ai-and-why-we-need-an-ai-specific-software-640f74aaf78f">Building AI software: Data-driven vs model-driven AI and why we need an AI-specific software…</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Chatbots have an identity problem. It’s time we got things straight.]]></title>
            <link>https://medium.com/@ronald_istos/chatbots-have-an-identity-problem-its-time-we-got-things-straight-fd0d3ac3fbb1?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/fd0d3ac3fbb1</guid>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[chatbot-development]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Mon, 25 Jun 2018 22:35:12 GMT</pubDate>
            <atom:updated>2020-01-15T09:03:02.958Z</atom:updated>
            <content:encoded><![CDATA[<h4>Let us stop talking about machine learning, neural networks, natural language processing and all the fancy tech. We need to be clear about what chatbots are and what they are not.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*etFYnLBxXbl5srgi" /></figure><p>It is entirely our fault. As technologists we got too carried away with the technological potential. Two minutes into a “why chatbots” conversation and we are digging a hole for ourselves trying to explain natural language processing, artificial intelligence and all the other <em>really</em> exciting things.</p><p>We then make an even worse mistake and pitch chatbots against other technologies. 2016 was all about how <a href="https://chatbotsmagazine.com/how-bots-will-completely-kill-websites-and-mobile-apps-656db8e6fc03">chatbots were going to replace the web</a> and how they <a href="https://www.telegraph.co.uk/technology/2016/03/31/the-end-of-apps-is-here-long-live-chat-bots/">were going to replace apps</a>.</p><p>Things didn’t quite work out that way.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1YliZ6jmO2WwKfEL" /><figcaption>Chatbots need to eat some humble pie.</figcaption></figure><p>Is that because chatbots are not a useful technology?</p><p>Thankfully not.</p><blockquote>Chatbots can be extremely useful by doing very little and often work best when used in conjunction with other technologies like websites or apps. That, however, is not that exciting to talk about.</blockquote><p><a href="https://blog.intercom.com/simple-question-bot-asks-lead-qualification/">This post by Intercom is a great example</a> of not getting carried away. They admit that one of the most useful things the <a href="https://hackernoon.com/tagged/chatbot">chatbot</a> does is categorise contacts into either exiting customers or new customers. Based on that, users were routed to the right team. Problem solved, money saved. No AI involved.</p><h3>Someone is wrong on the web.</h3><p>Try to search “what is a chatbot” on <a href="https://hackernoon.com/tagged/google">Google</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*M77It0rxe-FBsWyrfHksMg.png" /><figcaption>“what is a chatbot” search results</figcaption></figure><p>Google starts out with two definitions and both are highly problematic.</p><p>The dictionary definition says chatbots are “designed to simulate conversation with human users”. That makes it sound like chatbots are supposed to sound human-like or be indistinguishable from a human.</p><p>The Wikipedia definition is even worse. A chatbot is a computer program <em>or</em> an artificial intelligence. As if those two things are somehow separate.</p><p>Scroll down the page and most results are a mismatch of confusing chatbots with the technology one could use to develop chatbots (e.g. natural language processing or the even more enigmatic AI) or a problem chatbots are meant to solve (e.g. customer support).</p><p>To give you an idea of what clarity looks like let us search for the definition of a website.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*og5RebPxMa2RrA4LgFO8ug.png" /><figcaption>“what is a website” search results</figcaption></figure><p>Even scrolling down the page all you get are similar definitions. No mention of HTML, CSS, Javascript or ecommerce or education or any specific issue.<em> There is no confusion between the end result and the tools to create that or the purpose it serves.</em></p><p>We need to talk about chatbots in the same way.</p><blockquote>Chatbots are computer programs that interact with a user using a conversational paradigm.</blockquote><p>That’s it. Everything else comes afterwards.</p><p>This is significant because it does away with a lot of the noise and it simplifies the discussion about the utility of chatbots.</p><p>Is a conversational paradigm suitable? The answers could be:</p><ol><li>Yes, because it offers another channel to communicate with our users (e.g. Messenger)</li><li>Yes, because it is the easiest way to solve our user’s problem (e.g. customer support)</li><li>No, because the task at hand is best solved through a “classic” web interface.</li></ol><p>and so on and so forth.</p><p>How you will then build a chatbot, whether you should use NLP or any other fancy technology (<a href="https://chatbotsmagazine.com/chatbots-and-blockchain-are-a-perfect-match-99760ca4bcf8">blockchain anyone</a>?)is an entirely different issue.</p><p>Let’s start talking about chatbots as potentially very simple things that can do potentially very useful things. Let’s stop talking about AI, machine learning and all the rest until they are actually involved in the process in a significant way.</p><p>At <a href="https://greenshootlabs.com/">GreenShoot Labs</a> we help organisations define and develop chatbot-centric solutions.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fd0d3ac3fbb1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The future of work]]></title>
            <link>https://medium.com/@ronald_istos/the-future-of-work-6fb3cd83f5b8?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/6fb3cd83f5b8</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[bots]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Wed, 20 Jun 2018 14:12:30 GMT</pubDate>
            <atom:updated>2018-06-21T10:12:19.062Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*nUnNdHP7Fd4xbCA8.jpg" /><figcaption>Image courtesy of <a href="http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/system360/impacts/">IBM historical archive</a></figcaption></figure><blockquote>“We need to invest in a computer. Period. We need to tell our clients we are thinking about the future.”</blockquote><p>Jim Cutler, a character in the TV series Mad Men, had this epiphany to share with his colleagues. He, like many others in the late 1960s, was realising that computers were about to change how work got done.</p><p>When the computer finally arrives at SC&amp;P, the creatives of the agency are kicked out of their space to install the new behemoth. Technology replacing humans in the most literal way. One of the creatives, Ginsberg, actually has a nervous breakdown and mutilates himself. In a bizarre turn of events Ginsberg cuts his own nipple out.</p><h4>The future of work is already here</h4><blockquote>Digital transformation is no new thing. It has been happening for decades.</blockquote><p>Every wave of transformation has its own characteristics based on the technologies involved. What remains constant is that key to the actual change is a change in the “hearts and minds” of people that will be using the technology and adapting to new processes. Understanding the driving forces behind the current digital transformation wave and learning how best to make use of new opportunities (without causing anyone any nervous breakdowns) is our goal at Future Work and <a href="https://greenshootlabs.com/">GreenShoot Labs</a>.</p><p>In particular, <strong>we focus on the cross-section between conversational interfaces and automation</strong>. Digital transformation has multiple facets and ways to be done. We think there is something specifically interesting to do with conversation, chatbots and how they are becoming our interface into automated services and access to data.</p><h3>Conversational Interfaces</h3><p>There is a new breed of tool and it is taking over business everywhere. The main ingredients, as with any popular recipe, are simple:</p><ol><li>A means for teams to chat through any device.</li><li>Subdivisions in channels (or rooms or groups) to focus on specific subjects.</li><li>Some form of team directory with associated (although not crucial) presence indicator.</li><li>Integration into other systems via APIs and exposure of that functionality through chatbots.</li></ol><p>Slack is perhaps the most recognisable one, but it is by no means the declared winner. Just looking at the landing pages of some of the main players tells us a lot, check out the bold statements for each one:</p><p><strong><em>Slack</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QEPhbcv1ohsVMNegpO-kqw.png" /><figcaption>Slack Website</figcaption></figure><p><strong><em>Stride</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PHGqL9pqrp24jrMqzgsL-Q.png" /><figcaption>Stride Website</figcaption></figure><p><strong><em>Workplace by Facebook</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lRf_eb-QSUZ7D2GT4B3HVw.png" /><figcaption>Workplace by Facebook</figcaption></figure><p><strong><em>Microsoft Teams</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DklL3lWroIesw0yb-POf_g.png" /><figcaption>Microsfot Teams</figcaption></figure><p>Slack states it most boldly. It is “where work happens”. We agree. That really embraces the concept. It is not a tool like any other tools. It is the actual interface to work.</p><h3>Data and Automation</h3><p>Communication tools for teams have been around for many years. Skype is fifteen years old. What is different about these new tools is, on the one hand, their ability to easily capture the fluid structure and relationships of a team and, on the other hand, their ability to integrate with all the other tools we use.</p><blockquote>The ability to pull into the context of conversations the data that enables decisions to be informed and to push out of the context of conversations actions that get carried out by other tools is revolutionary.</blockquote><h4>Pulling in</h4><p>Say you are having a discussion about next steps for marketing actions. A marketing data bot can be queried to inform the debate.</p><p>“How many conversions did we get from our last campaign?”</p><p>It is easy to underestimate how disrupting and currently hard for most organisations it is to actually inform their conversations in such a fluid way.</p><p>Conversational interfaces, bots and data allow us to augment our conversations in a more natural and fluid way. Just like Picard and his team on the deck of the Enterprise we can invoke the “Computer” to inform our debate of whether to fight or flee from that Borg ship.</p><h4>Pushing Out</h4><p>Similarly, we can push out work from conversational interfaces. Stride tackles this head on by allowing us to highlight decisions from conversations and pull them out of the flow.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NC3Wg1CRnXxI5seQeymu5g.png" /><figcaption>Stride has native support for tasks and decisions.</figcaption></figure><p>More generally though, bots allow us to perform actions that would otherwise require a complete context change, remembering how a different interface works, etc.</p><h3>What’s next</h3><p>We are excited about how conversational interfaces, bots and automation are going to change the workplace. <a href="https://greenshootlabs.com/">We are working</a> to make it a more efficient, friendly, creative and ultimately human place. The tools are here to augment our capabilities and free up our time to do better, more creative work and to be at work less. We are looking forward to discussing and sharing findings.</p><p>Follow this magazine for news and updates, join our <a href="https://www.meetup.com/Future-Work-Bots-Automation-and-Digital-Transformation/">meetup</a> (we are starting in London).</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6fb3cd83f5b8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using a chatbot to grow awareness and train your team about GDPR]]></title>
            <link>https://medium.com/chat-bots/using-a-chatbot-to-grow-awareness-and-train-your-team-about-gdpr-4e69326a5f58?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/4e69326a5f58</guid>
            <category><![CDATA[slack]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[training]]></category>
            <category><![CDATA[gdpr]]></category>
            <category><![CDATA[chatbots]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Wed, 09 May 2018 04:59:02 GMT</pubDate>
            <atom:updated>2018-06-25T05:24:44.791Z</atom:updated>
            <content:encoded><![CDATA[<h4>To explore how training can be delivered through conversational interfaces we built a free Slack bot to raise GDPR awareness for any team.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yWQpKfrVx4gq8nl-pzCJzA.jpeg" /></figure><p>An area of focus at <a href="https://www.deeson.co.uk/">Deeson</a> is exploring how collaborative messaging environments like Slack can improve the way we can work. The soon-to-be-introduced set of EU regulations around privacy and data protection (<a href="https://www.deeson.co.uk/blog/gdpr-guide-you-wont-hate">GDPR</a>) provided with a useful way to test an idea we’ve had around delivering training through Slack.</p><p>A big part of successfully introducing GDPR is raising awareness within the entire organisation. Letting everyone know what the new regulations mean and how the entire team should be more conscious of what constitutes user data and how it is used.</p><h3>Why use chatbots to deliver training?</h3><p>Our hypothesis is that a chatbot can:</p><p>- be an effective way to deliver training without having to introduce additional applications, web interfaces and in general complexity for the user.</p><p>- take advantage of the full context of the team as provided by Slack so there is no further configuration required.</p><p>- it can keep track of what stage of the training people are at so they can pick up or stop at any time.</p><p>- can introduce a game / learning element through a quiz.</p><p>In short, Slack can provide a very efficient environment that can be exploited by a pro-active bot to deliver training and keep track of the overall status with minimal setup required.</p><h3>Designing and Building the GDPR Team Training Bot (in a week)</h3><p>Given our previous work with Slack bots and our product work (<a href="https://teamassist.ai/">TeamChecklist</a>) we were well placed to get a quick start. The bot specification we came up with dictated that when a user adds the bot to a Slack team they will:</p><ol><li>Be invited to take the training / quiz themselves.</li><li>Be able to invite specific individuals or the entire team to take the training and quiz.</li><li>Have a way to check on overall progress through anonymised aggregate quiz results.</li></ol><p>The way people are invited to the training is by the bot itself “introducing itself” to a person and asking them if they want to start with the training or the quiz.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rc0Rh7R0IVwHcoEsecZwrg.png" /></figure><p>The training itself is design as a set of interconnected modules with one module leading on to the next.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ekw7IBYk080y8vA9GcbcaQ.png" /></figure><p>The bot keeps track of context by following at which stage of the training each user is and what is their score on the quiz. If you answer wrongly the bot will provide feedback so that you can pick up the correct concepts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2Y5gTeS2559RayomIHwFeQ.png" /></figure><p>That’s it! We specified the bot and built it in a week.</p><h3>Deploying through the Slack app store</h3><p>The <a href="https://slack.com/apps/AADS3UMPT-gdpr-team-training">bot is now live on the Slack app</a> store and has it’s own landing page and related privacy policy information <a href="https://www.deeson.co.uk/gdpr-bot">here</a>. It’s a super simple way to raise GDPR awareness amongst the entire team and it is free for everyone. We will keep refining the information and behaviour itself based on your feedback so please let us know what you think.</p><p><em>At GreenShoot Labs we explore how conversational interfaces and AI can provide new ways to do work and engage with users and help our clients build </em><a href="https://greenshootlabs.com/"><em>chatbots</em></a><em>. We also build useful bots for teams, like </em><a href="https://teamassist.ai/team-checklist"><em>TeamChecklist</em></a><em> — a bot for managing automated tasks and reminders for the whole team in Slack.</em></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2F58521e%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2F58521e%2F&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c90883b0516ddbe5c59960c12a04bc64/href">https://medium.com/media/c90883b0516ddbe5c59960c12a04bc64/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4e69326a5f58" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/using-a-chatbot-to-grow-awareness-and-train-your-team-about-gdpr-4e69326a5f58">Using a chatbot to grow awareness and train your team about GDPR</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The most important metric for new SaaS products – Raw Product Ep.3]]></title>
            <link>https://medium.com/hackernoon/the-most-important-metric-for-new-saas-products-raw-product-ep-3-80a2904607b9?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/80a2904607b9</guid>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[growth]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Fri, 13 Apr 2018 17:32:41 GMT</pubDate>
            <atom:updated>2018-04-13T17:36:37.522Z</atom:updated>
            <content:encoded><![CDATA[<h4>At the start of a product’s life there is just one metric that counts and only you can define it. In this episode of Raw Product Italk about how you need to avoid distracting metrics and go straight to the heart of the matter.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kiLxH8DR_A69wllE." /><figcaption>Photo by <a href="https://unsplash.com/@markusspiske?utm_source=medium&amp;utm_medium=referral">Markus Spiske</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>So far in the “Raw Product” series we talked about how we <a href="https://hackernoon.com/raw-product-episode-1-launching-on-the-slack-directory-and-how-we-doubled-signups-with-a-simple-48f9c71f7c53">launched</a> TeamChecklist on the Slack app directory and how we <a href="https://hackernoon.com/how-we-build-our-saas-product-in-1-week-sprints-raw-product-ep-2-4e9f6cf78b02">build our product in weekly sprints</a>. This time round we turn our attention to metrics — specifically the most important metric early on for our SaaS product.</p><p>There is a wealth of information on how to measure the performance of a SaaS business. Metrics like monthly recurring revenue, cost of acquisition, lifetime value, churn rates need to become part of a SaaS business owners vocabulary. These metrics are useful in capturing and condensing the state of your business and providing indications about what you should be fixing. However, when you are just starting these metrics can feel a bit pointless. That’s ok.</p><p>After all what is the point of measuring monthly recurring revenue if it is in the single digits? If it won’t even buy me dinner? Cost of acquisition and lifetime value? I haven’t even worked out how I am going to get to a hundred clients!</p><p>If you are growing slowly and organically, as we are, in the very first months some of the standard SaaS metrics are ultimately pointless. You only have a few users, the product is in constant flux and pricing is still very much an experiment.</p><p>If you are adding many users quickly with a big marketing budget and a carefully orchestrated growth campaign the numbers may be more interesting to follow. However, you are still in a state of constant flux as you still need to better understand your product and how changes to one aspect will affect others.</p><p>So what does this mean? Should you stop measuring?</p><p>Of course not. We love to measure things!</p><p><strong>You need to measure the right things for each stage of the process.</strong></p><p>What we are measuring in these early stages is the <strong>utility of the product for our customers.</strong></p><blockquote><strong>We are trying to figure out just how much they love the product.</strong></blockquote><p>Nobody can tell you exactly what is the right measurement of that for your product. It is not a generic SaaS metric. You are going to need to find it out and you have to be brutal in choosing it.</p><p>Sometimes we choose the metrics that we know will please us:</p><p><em>Look at all those page views!</em></p><p><em>We had a thousand new sign-ups yesterday!</em></p><p><em>Our product is super sticky — they are spending so much time in the app!</em></p><p>However, it is not always page views, or sign-ups or generic engagement that matters.</p><p>Are users busy and active on your application because they getting value out of it or are they just clicking around trying to figure stuff out? Are you getting lots of sign-ups because your marketing is good or your app?</p><p><strong>Ask yourselves — what jobs is my tool supposed to help people fulfil? How do I know if that job is done? That’s what you should be measuring.</strong></p><p>In the case of <a href="https://teamassist.ai/team-checklist/"><strong>TeamChecklist</strong></a> the job is that of helping people capture their key processes in checklists and easily organise with their teams in Slack to get the tasks done. What we are measuring is a function of how many checklists people are creating, how many tasks those checklists have, how many tasks are actually completed and how many people interact with those checklists. We are building a model and refining the weights we apply to different actions but we are only measuring these real actions.</p><p>Page views, signups, MRRs are there but they are not the most important metric right now. Whether your product is an actually useful tool is the most important metric at the start. The rest will follow.</p><p><em>What is the key metric of utility for your tool? Are you measuring that?</em></p><p><a href="https://greenshootlabs.com/"><em>GreenShoot Labs</em></a><em> builds products that use automation, data and conversational interfaces to make teams happier and more productive.</em></p><p><em>Our first product is live on the Slack App Directory — </em><a href="https://slack.com/apps/A8N5U44CA-teamchecklist-tasks-and-reminders"><em>TeamChecklist</em></a><em> makes working through common, repeatable processes and tasks better for your team with automated notifications and reporting.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=80a2904607b9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/the-most-important-metric-for-new-saas-products-raw-product-ep-3-80a2904607b9">The most important metric for new SaaS products – Raw Product Ep.3</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What if Facebook were the cure for cancer?]]></title>
            <link>https://medium.com/@ronald_istos/what-if-facebook-were-the-cure-for-cancer-e2553b497cd9?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/e2553b497cd9</guid>
            <category><![CDATA[politics]]></category>
            <category><![CDATA[cambridge-analytica]]></category>
            <category><![CDATA[facebook]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[democracy]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Tue, 27 Mar 2018 17:35:23 GMT</pubDate>
            <atom:updated>2018-03-27T17:59:36.795Z</atom:updated>
            <content:encoded><![CDATA[<h4>It seems like we are finally questioning, in a meaningful manner, what influence companies like Facebook hold over society and what obligations derive as a result of that. We should not be afraid to demand more regulation.</h4><p>Society’s concern about tech companies is reaching a critical point. We can no longer pretend that it is something to <em>eventually</em> deal with sometime in the future. Quite literally <a href="https://www.theguardian.com/news/series/cambridge-analytica-files">democracy in countries across the world depends on us dealing with this issue now</a>.</p><p>For years we’ve heard the reasons we <em>shouldn’t</em> be asking for regulation. These messages have been drummed into us:</p><ol><li>This is a free market society — let the consumers decide.</li><li>The product is free so you are the product. Just accept it.</li><li>Privacy no longer exists.</li><li>Government-backed regulation is inherently broken — to avoid at all costs. Let them naturally self-regulate.</li><li>Technology companies are doing so much good. We need to allow them to <em>move fast and break things</em> — a phrase that comes directly from Facebook culture.</li></ol><p>Here is another way to think of it.</p><p><strong>Let’s assume Facebook was literally, not figuratively, the cure for cancer</strong>.</p><p>A young Harvard drop-out shows up and says “Hey, I have the cure for cancer. I am offering it for free!” How would society treat that? Would we just swallow the pills and ask questions later?</p><p>A quick look at the <a href="https://www.fda.gov/Drugs/ResourcesForYou/Consumers/ucm289601.htm">US FDA site</a> or the <a href="http://www.ema.europa.eu/ema/index.jsp?curl=pages/about_us/general/general_content_000109.jsp&amp;mid=WC0b01ac0580028a47">EU Medicines Agency</a> reveals that if Facebook were actually the cure for cancer, it would have to go through a huge list of studies and preparation, at least a decade of work at great expense, before it could get to access our data.</p><blockquote>“New drug and device approval in the United States take an average of 12 and 7 years, respectively, from pre-clinical testing to approval. Costs for development of medical devices run into millions of dollars, and a recent study suggests that the entire cost for a new drug is in excess of $1 billion.” —<a href="https://www.sciencedirect.com/science/article/pii/S2452302X1600036X"> Drugs, Devices, and the FDA: Part 1: An Overview of Approval Processes for Drugs</a></blockquote><p>For the cure for cancer then, as a society, we fully anticipate that there is a huge burden of proof on the inventor to convince us that this is safe. This is because it is a <strong>big deal</strong> that carries with it <strong>risks for all of us</strong> so we regulate it.</p><p>We’ve hit the same point with large, global tech companies. Tech companies need to carry the burden of proof that they will do the right thing and be open to constant scrutiny and regulation.</p><blockquote>Facebook and the other large tech companies currently impact society as much as any drug for any cure. Arguably they influence society in a more fundamental way — they stir our thoughts which in turn direct our actions. Let us, at the very least, apply the same level of scrutiny we apply to medicine.</blockquote><p>Here is yet another way to of thinking of it:</p><blockquote>Tech companies have insider info on us, we should not be afraid to regulate it.</blockquote><p>Financial markets have the concept of insider trading.</p><blockquote><strong>Insider trading</strong> is the trading of a public company’s stock or other securities (such as bonds or stock options) by individuals with access to nonpublic information about the company. In various countries, some kinds of trading based on insider information is illegal. This is because it is seen as unfair to other investors who do not have access to the information, as the investor with insider information could potentially make larger profits than a typical investor could make. — Wikipedia</blockquote><p>This is precisely what Facebook and other tech companies have access to. Insider information on us. Our likes, dislikes, innermost fears and desires. They then use that data to target ads on us and change our opinions about the world.</p><p>They need to be accountable and they deserve regulation.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e2553b497cd9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How we build our SaaS product in 1-week sprints — Raw Product Ep. 2]]></title>
            <link>https://medium.com/hackernoon/how-we-build-our-saas-product-in-1-week-sprints-raw-product-ep-2-4e9f6cf78b02?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/4e9f6cf78b02</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[saas-product]]></category>
            <category><![CDATA[saas]]></category>
            <category><![CDATA[agile]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Wed, 21 Mar 2018 14:36:00 GMT</pubDate>
            <atom:updated>2018-03-21T14:36:00.721Z</atom:updated>
            <content:encoded><![CDATA[<h4>How we split our week at GreenShoot Labs and how that connects to our long-term goals and initiatives alongside some personal thoughts on the pros and cons of short 1-week sprints.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0EygjC_Zh8oqvNFLJlsxUA.jpeg" /></figure><p>In the <a href="https://hackernoon.com/raw-product-episode-1-launching-on-the-slack-directory-and-how-we-doubled-signups-with-a-simple-48f9c71f7c53">last instalment</a> of Raw Product we talked about our experience launching on the Slack store and some of the learnings from that. In this episode I share how we organize our week to try and combine long-term goals with more short-terms reactions in a hybrid Scrum/Kanban method.</p><p>The usual disclaimers apply. This is just how we do it — not THE way to do it and it works for our current team size (a small, focussed, distributed team of 6). I find reading how other teams approach these issues very useful so hopefully this is a way of paying forward for the knowledge and experience sharing that I gained from.</p><h3>The big picture — Goals and Initiatives</h3><p>Before diving into a single week it’s worth looking at the slightly bigger picture. Beyond our strategic vision we set ourselves monthly and quarterly goals with measurable outcomes. These then translate into specific initiatives (larger, coherent pieces of work) and an understanding of how much time we want to dedicate to any initiative.</p><p>For example, one of our current goals is to make the SaaS app buyable!</p><p>Yes, I know that is a pretty basic thing for a SaaS tool. We launched early and didn’t worry too much about how people would pay us. We threw together some somewhat reasonable prices, used <a href="https://spark.laravel.com/">Laravel Spark</a> and <a href="https://stripe.com/">Stripe</a> so we can accept payments but hardly touched any of the out-of-the-box settings. Now that we know people are actually interested in the tool and that they are even willing to go through the very basic payment process we are ready to invest the time to make that payment process enjoyable, more relevant and something that inspires confidence in our service. This means proper modelling of alternative price plans and implementation of the tech to support those plans, design that is in line with our current brand, and generally a process that reassures (it also means properly figuring out the joys of charging for VAT — sales tax — across the EU). These points also roughly translate into the initiatives that will support the overall goal of making the application buyable.</p><p>As a sidenote we have recently moved our planning into <a href="https://www.aha.io/">Aha!</a>. We find it a very comprehensive (even if somewhat complicated) tool and right now it fits very well with our approach. So much so that we have directly adopted from it a lot of its product terminology (things like goals, initiatives, ideas, features, etc).</p><p>Ok — with the big picture mostly in place lets look at how a single week shapes up.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lSQ7OcMNZuY-sWXGmPwfHg.jpeg" /></figure><h3>Monday</h3><h4>Release, analyse, re-plan (if needed)</h4><p>We try and release as often as possible but Monday is the “official” release day. Anything not released during the previous week that has passed all the code review, testing and QA phases will get released on Monday. The person responsible for the release has this as the main task for the day while others are available to jump in if issues arise — otherwise people get started with their sprint tasks. We hope one day to automate many more of these aspects so that release is not something someone has to be responsible for but we are not there yet.</p><p>Monday is also the day we analyse feedback from the previous week in a focussed way, and turn it into actions. We use tools like Google Analytics, Hotjar, New Relic and our own metrics on app engagement.</p><p>With data at hand, we then review our initiatives and decide whether any course correction is required. We ask ourselves questions such as:</p><ol><li>Are we seeing the expected outcomes from features that were released previously?</li><li>Is there any new data / understanding that points to different features to develop or a change in goals or initiatives?</li><li>Are there any infrastructure, security issues that require explicit attention?</li><li>From a technical and UX perspective, do any of the bigger pieces of work require re-thinking now that they are further along the way?</li></ol><p>Key to this being an effective process is that we are all very open and willing to accept that what we thought last week to be the absolutely best course of action may simply no longer be that.</p><blockquote><strong>Strong opinions loosely held</strong> is definitely emerging as a defining aspect of GreenShoot Labs culture.</blockquote><h3>Tuesday and Wednesday</h3><h4>Develop, design, mockup, discuss, plan</h4><p>Tuesday and Wednesday are getting stuff done days. It is a combination of deep, uninterrupted creating and collaboration where appropriate. That means coders code, designers design, UXers mockup, planners plan, and we all share ideas and support each other to get stuff done.</p><p>By Wednesday afternoon we would have ideally released a couple of new features, have mockups for new features, and have a plan for the week after next in terms of specific tasks to do.</p><h3>Thursday</h3><h4>Feedback</h4><p>By Thursday we aim to have enough done so that others can feedback on it. We can discuss technical challenges of features, and provide feedback on mockups. While not easy, our aim is to do this for the work for the week after next so that the thinking can have its time to brew in our individual brains.</p><h3>Friday</h3><h4>Wrap up, review, retrospective and plan next week</h4><p>By Friday we are wrapping up our sprint. We review work done, what we are going to release, what remains on the board, and what new items move into the board based on availability. We run a retrospective of the week, identify what went well, what didn’t and what we will aim to improve.</p><h3>The good, the bad and the ugly</h3><p>The weekly map I describe is a somewhat idealised version. We try to stick to it as closely as possible but we don’t beat ourselves up if it doesn’t always go like that. We are a small team and we need to deal with whatever issues come up. In general, here are some thoughts I have about the process:</p><ol><li><strong>1-week sprints saved us multiple times. </strong>We are a small team, we communicate frequently. Still we’ve found that things can go off course. Had we not had 1-week sprints those deviations would have been actually costly. 1-week sprints force all the units of work and planning into far more maneageable pieces.</li><li><strong>I am not sure how it would scale to bigger teams.</strong> Before GreenShoot Labs I almost exclusively worked in 2 or 3 week sprints. Also a lot of the work was from within an agency for customers. I dread to think of having to run at this pace with a large team with multiple stakeholders. Having said that I would very reluctantly go back to bigger sprints.</li><li><strong>The pace is exhausting and fantastic. </strong>The running joke is that by the time we hit the end of the week someone will invariable pop-up and say “I can’t believe it is Friday already”. I love that.</li><li><strong>Balance with the rest of life is crucial. </strong>To be able to keep up the pace we need to … pace ourselves. A few months back I went through a few weeks of much more intense work in order to get stuff built but quickly scaled back. It was not sustainable.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ofYz-C_W7Tom8qiD." /></figure><p>My (obvious but not so easy to implement) learning:</p><blockquote>Building a SaaS company (especially a bootstrapped one) is a marathon. Perseverance and continuity is far more important than burning out in a ball of fire.</blockquote><p>That’s our week. Would love to hear your thoughts, suggestions and if you have a similar post please add it to the comments.</p><p><a href="https://greenshootlabs.com/">GreenShoot Labs</a> builds products that use automation, data and conversational interfaces to make teams happier and more productive.</p><p>Our first product is live on the Slack App Directory — <a href="https://slack.com/apps/A8N5U44CA-teamchecklist-tasks-and-reminders">TeamChecklist</a> makes working through common, repeatable processes and tasks better for your team with automated notifications and reporting.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4e9f6cf78b02" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/how-we-build-our-saas-product-in-1-week-sprints-raw-product-ep-2-4e9f6cf78b02">How we build our SaaS product in 1-week sprints — Raw Product Ep. 2</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Raw Product — Episode 1: Launching on the Slack Directory and how we doubled signups with a simple…]]></title>
            <link>https://medium.com/hackernoon/raw-product-episode-1-launching-on-the-slack-directory-and-how-we-doubled-signups-with-a-simple-48f9c71f7c53?source=rss-5ca576794be4------2</link>
            <guid isPermaLink="false">https://medium.com/p/48f9c71f7c53</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[slack]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Ronald Ashri]]></dc:creator>
            <pubDate>Thu, 08 Mar 2018 16:01:12 GMT</pubDate>
            <atom:updated>2018-03-10T07:44:52.334Z</atom:updated>
            <content:encoded><![CDATA[<h3>Raw Product — Episode 1: Launching on the Slack Directory and how we doubled signups with a simple change</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Wgh0X1C4ETcMUpzuKBwqzA.png" /></figure><p><em>This is the first in a series of posts called “Raw Product”, sharing our experiences of building products for conversational platforms. The title captures what we are doing at GreenShoot Labs — minimum fuss, minimum fluff (most of the time) — simply focused on building useful products. For Episode 1 we share some of the early learnings of our first product release on the Slack Directory and how we doubled sign-ups with a simple change.</em></p><h3>Prologue: launching GreenShoot Labs</h3><p>We started discussing the possibility of building products with <a href="https://medium.com/u/17d031fe91cf">Tim Deeson</a> around April 2017. Tim is the founder of the <a href="https://deeson.co.uk/">Deeson</a> digital agency, and I was working there on technical strategy. We shared a common interest in product development, conversational interfaces (chatbots) and artificial intelligence; especially when these technologies are used in service of supporting teams to work better together. We joined the dots, and <a href="https://greenshootlabs.com/">GreenShoot Labs</a> was born — a product studio closely associated to Deeson and focussed on building products that make teams happier and more productive.</p><p>Over the next couple of months, as we worked out the details of how GreenShoot Labs would operate, lofty ideas were exchanged about chatbots, AI, machine learning and the future of work. Comprehensive Google docs were created to define our strategy and compliments were received from friends, colleagues and advisors on the breadth and depth of our vision.</p><p><em>Then we stopped messing about and started building an actual product.</em></p><h3>Deciding on a product idea</h3><p>We have several product ideas, ranging from very simple to very complex. Ideas, after all, are cheap. As a first try, however, we picked something we thought was on the simpler side of the scale. A checklist tool designed for Slack teams.</p><p>Tim wrote <a href="https://medium.com/teamassist/introducing-teamchecklist-share-best-practice-using-checklists-in-slack-1839bff9afbf">here</a> about why we created TeamChecklist, why checklists are useful for teams in general and how they can be made even more useful within a conversational platform.</p><p>Following classic lean startup lessons, we started by creating a landing page and running some ad campaigns to see the amount of interest the tool would get. While we had some validation, we realised that t<em>argeting just Slack users that knew about, were interested in and were actively looking for bots as a way to enhance their Slack workspace was more of a marketing exercise than what we could reasonably handle</em>.</p><blockquote>In fact, the majority of Slack users we talked to had only a minimal interest in bots and most didn’t even know a Slack app directory existed!</blockquote><p>However, we still knew that people do use bots. It is clear this is a growing market, and the danger of entering early(ish) is that you will not easily find lots of users through traditional marketing techniques.</p><blockquote>We decided that the simplest way to do it was to put our product on the Slack marketplace shelves and see if people would pick it up.</blockquote><p>We thought that from an engineering perspective it would be a simple enough challenge and we were rich in engineering resource. The exercise would allow us to learn the Slack API and build a SaaS platform without getting too distracted in the complexity of the tool itself (we later found out we were wrong, but that is a different post!).</p><h3>Submitting TeamChecklist to the Slack Directory</h3><p>Work started in earnest in October 2017, and by January 2018 we launched on the Slack directory.</p><p>The Slack submission process was smooth. We <a href="https://api.slack.com/docs/slack-apps-guidelines">read the guidelines</a> and set up a checklist (of course!) with all the requirements. Once we were happy that we met all of them we hit the submit button to get the application reviewed.</p><p>Slack was pleasantly reactive. We saw that someone logged in to the application and the next day we got a set of questions about the <em>permissions</em> we were requesting regarding how much of a Slack workspace’s data and functionality our application should access. Slack rightly pointed out that we were asking for permissions we didn’t seem to be using or need. We updated our application, re-submitted and after another review, we were good to go!</p><p><a href="https://slack.com/apps/A8N5U44CA-teamchecklist-tasks-and-reminders">TeamChecklist</a> was finally on the Slack Directory!</p><p>We then went to test a sign-up from the store and… it didn’t work! Doh! Turns out that there was still one place in our application where we were requesting permissions that we were not permitted to use as a live app. This is not an error that would show up in the test app. A quick patch later and we were finally good to go.</p><h3>Users!</h3><p>The first user signed-up a few hours after we went live. We were genuinely surprised. Our app was at the very bottom of the categories we submitted it for (Office Management, Productivity, Project Management) and we hadn’t done any marketing. How were people finding us?</p><p>Over the next few days, we saw a steady stream of user sign-ups. Our own marketing site was getting zero visits; our <a href="https://www.youtube.com/watch?v=euxS04SbbpI">introductory video</a> was getting only a fraction of views compared to sign-ups. This was all about the Slack directory. People could install straight from the directory and they didn’t look for any further information.</p><blockquote>Our initial hypothesis of the most efficient test being that of putting the tool on the Slack directory and seeing who picked it was right.</blockquote><p>With the tool on the directory we skipped all the marketing problems of narrowing it down to people that know about Slack bots and are motivated enough to install them. We got a clear signal regarding the specific idea with just our target market in play.</p><p>It was as simple as people looking for a checklist tool and the title of our tool — TeamChecklist — fitting that description. Users were typing “checklist” in the search box and we showed up as an option.</p><h3>Iterating on the value proposition</h3><p>Now, given that learning, we also realised that we had to take a closer look at our title and the short description of the tool. Initially, it was:</p><p><strong>TeamChecklist: help your team share best practice with reusable checklists.</strong></p><p>Our value proposition was focussed on how checklists were a great way to capture a team’s set of best practices and that TeamChecklist would allow you to then share and reuse those processes.</p><p>Analysing the <em>actual </em>usage of the tool we realised that users had a broader set of scenarios in mind. While some were interested in best practices most were just looking for a checklist tool that was tightly integrated with Slack. So we formulated the hypothesis that if the value proposition was less about best practices and more about the raw functionality we should get more signups. In interviews with a handful of users, they all also agreed that the value proposition “share best practices” was either too confusing or too abstract.</p><p>We changed our listing title on the store to:</p><p><strong>TeamChecklist — Tasks and Reminders: Create checklists, assign to anyone in your team — receive automated task reminders and daily reports.</strong></p><p>Pure focus on just what the tool did. We also tightened up the rest of the copy to be much more specific and practical.</p><p>We submitted and the next day got approved to take our updated listing live.</p><blockquote>We saw the rate of signups double almost immediately and it has remained consistently so.</blockquote><h3>Summary</h3><p>So our first set of lessons is:</p><ol><li>There is a large number of people interested in Slack applications — however, marketing to them through traditional means is still hard. Listing on the Slack directory may be the most efficient way to get to them.</li><li>Most people don’t look at anything other than the actual listing — make sure it fits your users’ expectations and be willing to experiment. We could skip the website for the app, and we would still be getting signups — people are not visiting it. The listing is that much more critical for us.</li><li>Allow people to install and try the app directly from the directory. Reduce all friction between the discovery of the app and use of the app.</li></ol><h3>Next Up</h3><p>We are now focussing on increasing engagement within the application. We want to get better at measuring the usefulness of the app and figure out what features that will help improve it. In the next instalment of Raw Product, I will talk about our product development process and the metrics we are putting in place to measure engagement. In the meantime, if you have any questions about the Slack store submission process or what GreenShoot is working on hit me up on <a href="https://twitter.com/ronald_istos">Twitter</a>.</p><blockquote>Follow <a href="https://twitter.com/teamassist_ai">TeamAssist</a> for news about TeamChecklist and checkout the tool on the <a href="https://slack.com/apps/A8N5U44CA-teamchecklist-tasks-and-reminders">Slack Directory</a>.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48f9c71f7c53" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/raw-product-episode-1-launching-on-the-slack-directory-and-how-we-doubled-signups-with-a-simple-48f9c71f7c53">Raw Product — Episode 1: Launching on the Slack Directory and how we doubled signups with a simple…</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>