<?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 Eddy Wong on Medium]]></title>
        <description><![CDATA[Stories by Eddy Wong on Medium]]></description>
        <link>https://medium.com/@eddywongch?source=rss-90e3eb1bdb60------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*b1S-09-TRX9lUUKM.jpeg</url>
            <title>Stories by Eddy Wong on Medium</title>
            <link>https://medium.com/@eddywongch?source=rss-90e3eb1bdb60------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 00:26:56 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@eddywongch/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[The Lost Story of Open Sesame, an AI Startup of the Dot-Com Era]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@eddywongch/the-lost-story-of-open-sesame-an-ai-startup-of-the-dot-com-era-9a7e8b0adcb9?source=rss-90e3eb1bdb60------2"><img src="https://cdn-images-1.medium.com/max/600/1*dTFTTtsT28i8zKLZXP8T6g.jpeg" width="600"></a></p><p class="medium-feed-snippet">Long before &#x201C;AI Agent&#x201D; became the buzz word of the day, a small Cambridge, Massachusetts research shop built an AI agent &#x2014; and shipped it&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@eddywongch/the-lost-story-of-open-sesame-an-ai-startup-of-the-dot-com-era-9a7e8b0adcb9?source=rss-90e3eb1bdb60------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@eddywongch/the-lost-story-of-open-sesame-an-ai-startup-of-the-dot-com-era-9a7e8b0adcb9?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/9a7e8b0adcb9</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[personalization]]></category>
            <category><![CDATA[agents]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[dot-com]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Tue, 04 Aug 2026 17:48:03 GMT</pubDate>
            <atom:updated>2026-08-04T17:48:03.811Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Cassandra/DataStax Community Resources]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@eddywongch/cassandra-datastax-community-resources-ef1cfb01223b?source=rss-90e3eb1bdb60------2"><img src="https://cdn-images-1.medium.com/max/1600/1*2QUhprzVwQ2s4Q4uRuFh2w.png" width="1600"></a></p><p class="medium-feed-snippet">Books</p><p class="medium-feed-link"><a href="https://medium.com/@eddywongch/cassandra-datastax-community-resources-ef1cfb01223b?source=rss-90e3eb1bdb60------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@eddywongch/cassandra-datastax-community-resources-ef1cfb01223b?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/ef1cfb01223b</guid>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Thu, 07 May 2020 19:18:04 GMT</pubDate>
            <atom:updated>2020-06-13T17:17:12.587Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Observations of the Peruvian Startup Ecosystem: Desperately Seeking CTOs]]></title>
            <link>https://medium.com/@eddywongch/observations-of-the-peruvian-startup-ecosystem-desperately-seeking-ctos-e73e22fa7f7a?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/e73e22fa7f7a</guid>
            <category><![CDATA[education]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[peru]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Wed, 05 Dec 2018 23:48:36 GMT</pubDate>
            <atom:updated>2018-12-05T23:48:36.141Z</atom:updated>
            <content:encoded><![CDATA[<p>At the Venture Capital Conference in April of 2016 in Lima, Peru, one of the panelists shouted, “It’s very hard to find a CTO in Peru.” As EIR (Entrepreneur in Residence) in the Peruvian ecosystem and as judge at <a href="http://www.start-up.pe/">StartupPeru</a> (the national government accelerator), I had the chance to listen to more than 200 startups in Lima, Arequipa and Piura. Many of them displayed this same problem. Startups have trouble finding CTOs, and it seemed to be more the rule than the exception.</p><p>A high number of Peruvian startups have their beginnings without a technical co-founder, sometimes because they didn’t think they needed one (which is another different problem), but when they did, they settled for a junior person, mainly because “they could not find anybody else” or they could not or would not pay more. Why is this? I asked myself. Where are the experienced tech guys? Why aren’t tech folks founding or joining startups?</p><p>To find the answer I had to look at the past, to the 80s and 90s. A confluence of factors has conspired against the minting of “experienced” startup CTOs: improper education, a sui-generis IT corporate industry heavy on proprietary solutions and custom implementations, lack of experienced role models, the “brain drain” caused by attraction of opportunities abroad, plain economics and the mindset that these factors have created. In this post, I’ll share some of my own personal experiences, as I myself left Peru in the late 80s to study Computer Science (CS), settled in Boston (surrounded by tech), and eventually became CTO and cofounder of startups.</p><p><strong>The Lost Generation</strong></p><p>The 80s was terrible decade in Peruvian history. Although the country returned to democracy from dictatorship, it got mired in terrorism and hyper-inflation. Imports were very restricted and anything coming from abroad was difficult to get or extremely expensive. Once democracy was restored in 1980, imports started to trickle in.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TNgQ49G2LsxZ54hzioNkWA.jpeg" /><figcaption>CP Juan XXIII Computer Center flyer (in the 80s)</figcaption></figure><p>As a teenager I was lucky to attend a private school in Lima (CP Juan XXIII). It had superb infrastructure, relatively speaking, by Peruvian standards. I discovered my calling when the school acquired 20 Italian Olivetti M20 L1 personal computers (the school was run by Italian priests). The computers had monochrome green screens, 128K of RAM and stored data in 5 1/4 inch floppy disks. Dinosaurs by today’s standards. Soon after, the school started to offer courses in BASIC, a popular programming language at the time. While other kids asked their parents for money to go to the movies, I would ask my Mom for money to stay after school and take every computer course there was. During this time imports began to arrive in Peru. Arcade game parlors started to pop up. Locally, they were called “pinbol” (for pinball). Asteroids was one of my favorite video games and liked it so much that I wanted to replicate it. But, I found that language BASIC was too slow for a video game. However, in that attempt, I discovered by myself the true usefulness of trigonometry. I was learning more by attempting to animate ships than by listening to my math teacher. I was hungry for anything related to PCs, but information was difficult to obtain. There was no Internet back then, so computer magazines (like “Byte”) were my source of knowledge. They were 5 times the listed price and the only place where you could get them was at exclusive bookstore that sold imported, mostly in English-language books and magazines (“Librería ABC” in Miraflores). I grew up with the perception that anything that was imported was “good”, whereas anything local was second-rate. It became a habit to always seek out products with the label “made in USA” or “made in Germany.”</p><p>By my senior year in high school I was dreaming of going to a university Computer Science (CS) program. There were none in Peru. I wanted to learn the programming language “C”, so that I could write my Asteroids-like video game. It was a frustrating sensation not to have local options. There was no choice than to aim elsewhere, to the country that invented most computer and software technologies, the USA. So my sights were set to the US, as I had known about CS programs thru travel and thru relatives.</p><p>A “miracle” happened when I learned about the <a href="https://en.wikipedia.org/wiki/Fulbright_Program">Fulbright program</a>, a scholarship program to study in the US, sponsored by the US government. The director of the program paid a visit to my school one day and presented the availability of scholarships to study in the US. I promptly applied and after a rigorous and tedious process, I was in. I was one of lucky 15 applicants in the whole country that year. We were all sent to the US with undergraduate scholarships. I was going to study Computer Science at an American university, I couldn’t at first believe it. In my cohort, there were at least 2 other guys gunning also for CS. Given the situation the country was undergoing in the late 80s (terrorism, kidnaps), this was an extremely exciting opportunity, all I knew was that it was going to take me to a better place.</p><p>Years later, I finished college (Ohio Wesleyan University) and was accepted to grad school at Brown University, an Ivy League school and of one of the top 15 Computer Science programs in the US. After graduation, I was lucky to land a job in Boston and live thru very interesting eras in the computer, software and Internet industries. The Greater Boston area had long history of tech companies. It was the home of many legendary mini, PC, software and Internet companies like Digital Equipment, Honeywell, Prime Computer, Data General, Wang, Lotus, Allaire, ATG, Altavista. Boston became my home.</p><p>Many protagonists of the tech industry lived and worked down the street. I experienced first hand the PC era (rise of Apple-IBM-Microsoft-Intel-Lotus-3Com-Novell), the emergence of the Internet (the .edu extension just appeared while I was in school), the transition from mainframe to client-server (the fall of Digital Equipment), the rise of Web (W3C was down the street, and HTML and the browser were just invented), having an exit and a crash during the Dot-com era (Open Sesame), being an employee at startup that had an IPO (Cambridge Technology Partners) and so on.</p><p>I have heard of many friends and relatives leaving Peru for greener pastures, like me during that era. For over a decade or more, many left for the US, Canada, Chile, Venezuela, Brazil and Spain. This was a “lost generation” for Peru, I suspect that a whole swath of would-be IT professionals and computer scientists left in the decades of 80s, 90s and 2000s, leaving no legacy to the next generation. No role models, no mentors, no investors.</p><p><strong>Piracy, and the under valuation of software</strong></p><p>I visited Peru every couple of years, mainly to see family and friends, but always checking out progress and advances of the IT industry. One time around 2003, a friend of mine, who was a graphic designer, took to me to a marketplace called “Wilson” (for name the street “Av. Wilson”) where you can make all sorts of custom printing jobs. There you could find rows of little printing shops that can make business cards, letterheads, banners, you name it. It was also the place where you could get computer supplies and computer software. Yes, software, but the pirated one. Vendors would stand in front of their shops and offer you a latest copy of Adobe Photoshop for $1.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*R2l2Xk2jKcMgyGVCYkPkRg.jpeg" /><figcaption>Pirated Software for $1 a CD (late 90s and 2000s)</figcaption></figure><p>I look back to those years (late 90s to 2000s), and say to myself, that blatant piracy must have killed any hope for a market for software written locally. Lo and behold, many colleagues and friends told me that in those years anybody working in IT would be doing computer maintenance, custom services, software maintenance, or package implementation. That is, secondary occupations or professions. In other words, the focus was on using software (that came from abroad) rather than creating software. The sad reality was that a lot local university programs that served this market had the same orientation. It was the “management” of the usage of software or maintenance focus, rather than creation from scratch. Why? Because of software piracy, there was no market for software creation.</p><p><strong>The Wrong Mindset</strong></p><p>As an advisor to startups in the Peruvian ecosystem I see a pattern of structural mistakes that keeps repeating itself. They are essentially produced by the wrong mindset of the founding team and a flawed foundation of the startups. I would describe the mindset as heavily tilted (if not completely controlled) towards non-tech or commercial backgrounds and worrisomely with a lack of tech savviness. This mindset probably comes from vestiges of colonialism, where the “boss” (or owner) was the all-powerful and decision making head of the enterprise (most likely a plantation) and everybody else was a subordinate. This culture has been perpetuated by big enterprises managed by the few powerful families that controlled Peru for many decades. You notice this vestige particularly when you see a startup with no CTO, or a CTO as an employee. Technical roles are still regarded as subordinate roles and not as leadership or decision making role. This is completely opposite to the culture that you see in more advanced startup ecosystems (like Silicon Valley or Boston) where you could find CEOs with an MBA and tech backgrounds or techie CEOs altogether. In these advanced ecosystems a startup is valued even higher if the founding team or initial employees have more tech backgrounds. That reflection of valorization of tech knowledge has not occurred in Peru yet.</p><p>Another lacking trait that Peruvian founders suffer from is the multi-disciplinary background. On both sides. Business folks lack tech savviness and tech folks lack soft and business skills. This can be attributed to a rather rigid higher education system. Peruvian kids have to decide which university and which major in the middle of the senior year in high school, with no room for mistakes. Once you pick a program, you follow a prescriptive and static coursework or “malla curricular” that has very few electives and doesn’t allow “try” other fields. As a result, graduates are formed with a “siloed” mindset, that is, you are either an administrator, an accountant or an engineer and nothing in between. The corporate industry perpetuates this silo mentality by recruiting specific profiles. Employees are then assigned to heavily hierarchical departments.</p><p><strong>A New Hope</strong></p><p>So let’s fast forward to today and try to answer the original question: Why are there so few CTOs? One hypothesis could be, because a software industry never developed in Peru, software skills were not valued and still are not valued, and that the few talent that it had found better opportunities abroad. And because there was no native industry, there is no software culture or communities, therefore no role models or mentors that could nurture them.</p><p>In my opinion, if you want “mint” more people in Peru with a high probability to turn into technical leaders (like CTOs), you would want to improve these:</p><ul><li>Improve awareness of programing and <a href="https://en.wikipedia.org/wiki/Computational_thinking">computational thinking</a> at early age, not just for people that will go down the CS route, but for general aptitude.</li><li>Create higher education programs oriented towards “building” not just “managing” software.</li><li>Develop a market for people with software skills, the salary scale is tilted towards people with business skills.</li><li>Develop technical leadership roles, not just individual contributors.</li></ul><p><strong>A Brighter Future</strong></p><p>In summary, young technical people in Peru today have much a better chance of becoming software engineers, CTOs and startup founders. When that hunger for knowledge kicks in, when you are student, information and education are readily available on the Internet. This is completely different from what I faced when I was teenager. Here are some other reasons why there are better chances today (if I were to compare it to the kinds of opportunities that I had in 80s/90s):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*bS5D8QorgwfNsUBgro4HrA.jpeg" /><figcaption>Scratch programming in 5th grade at CP Juan XXIII today</figcaption></figure><ul><li>No restrictions on imports, the Internet makes knowledge transfer instantaneously.</li><li>Many initiatives in children education are popping up: coding school startups like <a href="http://silabuz.com/">Silabuz.com</a>, <a href="https://www.facebook.com/creacodeperu/">Creacode</a>, <a href="http://www.codenmicole.com/">Code en mi Cole</a> and <a href="https://www.crackthecode.pe/">Crack the Code</a> . My own alma mater Juan XXII is teaching Scratch in the 5th grade and Telefonica (a major phone, TV and Internet provider) has a “digital education bus” for kids operating under the <a href="http://www.fundaciontelefonica.com.pe/2018/11/14/mas-de-9-mil-escolares-de-lima-aprenderan-en-bus-educacion-con-tecnologia/">Fundación Telefónica</a> name.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/360/1*kt8ehbbedJu40oWVBBEwWw.jpeg" /><figcaption>Computer Science 101 at UTEC in Lima today, modeled after <a href="https://medium.com/@davidventuri/review-harvard-cs50-introduction-to-computer-science-75405e684a61">Harvard University’s CS50</a></figcaption></figure><ul><li>More and more accessible school and CS programs: Universidad Católica San Pablo (Arequipa), Universidad Nacional San Agustin (Arequipa), UTEC (Universidad de Ingeniería y Tecnología) are offering undergraduate computer science programs. UTEC’s intro to CS class is modeled after <a href="https://medium.com/@davidventuri/review-harvard-cs50-introduction-to-computer-science-75405e684a61">Harvard University’s CS50</a>, famous for being one of the best undergrad CS classes around the world. It’s required by all students, not just CS majors. San Agustin and San Pablo have a graduate schools in CS.</li><li>MOOCs offer many computer science courses online from top schools like Stanford and MIT. It is easier to connect with the online community with tech knowledge sites like Stackoverflow, Quora, and Stackshare.</li><li>No more piracy. Cloud and open source make it impossible to very hard to pirate software. Therefore there is a need for software talent, which creates a market for software skills.</li><li>You can make money in App stores. If you make an app that is universally useful or entertaining, there is global market for that and you can make a living, no matter where you live.</li><li>You can work remotely. Companies like <a href="https://www.upwork.com/">Upwork</a>, let’s you find freelance IT work and programming gigs no matter where you live.</li><li>Increasing local valorization of technology. CS is becoming mainstream. Just observe how many “technical” terms are heard in the popular culture: algorithms, AI, autonomous cars, data science, blockchain, Turing complete. There is global demand for computer science talent. In this field, there are no borders. This has not occurred in Peru yet, but market pressures should do their job, albeit slowly.</li><li>The new generation of Peruvian university students has a better sense of community. There are student-run organizations that promote entrepreneurship and agile processes. I anecdotally can cite two notable examples, <a href="https://web.facebook.com/cieutec/">Comunidad I&amp;E UTEC</a> and <a href="https://web.facebook.com/hultprizeatunjbg/">Hult Prize Chapter at UNJBG</a> (Universidad Nacional Jorge Basadre Grohmann), under the influence of American entities University Innovation Fellows and Hult Prize, respectively. I personally had met the (student) leaders of these organizations and I can say that they will get far.</li><li>The corporate mindset in Peru has still to change. CTOs are probably dead last in the terms of decision making authority and compensation. In the US, a curious change happened in the 80s going into the 90s. The Netscape IPO (1995) proved that a geek could be CEO and on top of the world. From then on, the sky was the limit for tech people. A catalyst like that has not happened in Peru yet.</li></ul><p>When I get question “how come it’s so difficult to find a CTO in Peru?”, my answer is multi-prong: it’s the education, the corporate industry and the history all at fault. There is no easy solution, than to work on improving the next generation of startup folks in order for them to have the right mindset and formation. The evidence that I have mentioned tells us that change is about to happen soon, but not tomorrow.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e73e22fa7f7a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Soft Skills of a Startup CTO]]></title>
            <link>https://medium.com/@eddywongch/the-soft-skills-of-a-startup-cto-ca94a16f23a7?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/ca94a16f23a7</guid>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[management]]></category>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Mon, 03 Sep 2018 14:15:18 GMT</pubDate>
            <atom:updated>2018-09-03T14:15:18.076Z</atom:updated>
            <content:encoded><![CDATA[<p>Recently a colleague of mine, Carlos Ganoza, posted an article about the “<a href="https://perustartup.blog/los-soft-skills-del-ceo-de-una-startup/">Soft Skills of a Startup CEO</a>.” In the same light, I thought of writing an article about the “<em>Soft Skills of a Startup CTO”</em>, since you would think that soft skills are associated with CEOs, but NOT with CTOs. Though CTOs are the ones who need them the most. Specifically, I wanted to share stories about the soft skills that I thought helped me in leading the tech efforts as CTO at a successful startup like <a href="https://www.wanderu.com/">Wanderu</a>.</p><p>My path to becoming a CTO was not a straight one. As most tech folks, I had a severe disadvantage: I was an over-protected introvert. Migrating to the US (from Peru) was my defining event, as I had to learn to fend for myself and become assertive and independent. Going from Lima, a big South American capital, to a small liberal arts college (Ohio Wesleyan University) in the “middle of nowhere, USA” was a good training ground for entrepreneurship, I believe. I had to figure out how to be resourceful and sometimes find creative ways to make up for the lack of resources, mine and that of my school. I had little spare money and my school was a bit behind on equipment. On one occasion, I volunteered to be a system administrator, so that I could get my hands on a Unix system and learn everything about it.</p><p>Following is a list of the soft skills that I believe helped me in carrying the job of CTO. They are what I think, the skills that every aspiring startup CTO should have.</p><p><strong>Communication</strong></p><p>Speaking in public in a second language was one the hardest barriers I had to conquer. Going to a small college where everyone knew each other and where everyone addressed each other by their first name, helped me with my self-confidence. I even stepped out of my comfort zone by becoming involved in the student-run activities (ACM — Association for Computing Machinery). This was my training ground for community-oriented events many years later.</p><p>In a startup situation, there are two main types of communication (by audience), technical and non-technical. Communicating with my fellow team-mate was the easy part. I liked explaining things on a whiteboard, so that came naturally. Explaining a design, a solution, a plan, is critical in order to move the troops in unison.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/403/1*0qMekuMpYLE6OXjqPYZ57w.jpeg" /><figcaption>Explainer in Chief</figcaption></figure><p>Communicating something technical to a non-technical audience, on the other hand, took extra effort and did not come easy. In order to be successful, one has to be self-aware and avoid the assumption that the audience knows what you are talking about. I had my practice working for a consulting company (<a href="https://en.wikipedia.org/wiki/Cambridge_Technology_Partners">Cambridge Technology Partners</a>), where I had to explain markup language concepts (<a href="https://en.wikipedia.org/wiki/Standard_Generalized_Markup_Language">SGML</a>, the predecessor of <a href="https://en.wikipedia.org/wiki/XML">XML</a>) like tags, elements, attributes and style sheets to writers who where English majors.</p><p>In a startup situation, you have to communicate to your partners, investors, customers and folks on the business side. Sometimes it was easier to come up with analogies or real life examples to express some technical concept. For example, I often mentioned the diagrams of routes on the back of airline magazines to explain what a <em>graph</em> is.</p><p><strong>Extroversion</strong></p><p>Most techies are introverted, that’s a fact of life. We like to be left alone in a corner and work on your stuff. This is detrimental if you are in a position of leadership. It won’t be of any use if you hide in corner to be by yourself.</p><p>It takes “effort” to be out there and chit-chat at a cocktail party or at a conference after-party. Being social is important because it can connect you to your next partner, investor, or employee. Networking and community formation are two important side skills evolving from this. If you are not somewhat extroverted or become more extroverted, you won’t succeed at forming a community. Navigating thru a community is nowadays critical in an open-source oriented startup software industry. It took some effort to get used to this, because I (like many techies) naturally find it exhausting.</p><p><strong>Leadership</strong></p><p>A CTO must chart the course and your fellow tech team-mates have to follow. Two of the most important tasks at the beginning of a startup is picking a tech stack and hiring your founding engineers. Picking a tech stack is one of the most nerve wracking things to do. What if I pick the wrong stack? I would find out many years later, if I did the right thing. In my experience, I tended to be bold in my decisions. I picked technologies that had sound concepts behind them and made sure that had a solid community was behind them. My examples were Java (server side) and XML in 1998 for a personalization engine (<a href="https://www.slideshare.net/slideshow/embed_code/key/b5x2wgvZAOQKzx">Open Sesame</a>) and nodejs, Python, MongoDB and Neo4j in 2012 for a ground transportation search engine (Wanderu). In the end, both picks turned up to be the excellent choices. One aspect of picking a tech stack is to attract fellow engineers. Most engineers like to work on interesting and leading edge tech. However, this has to be balanced with what the technology can accomplish and one has to be careful with the hype.</p><p>One of first things that a CTO will do at a startup is to hire a tech team. Hiring people whom you have worked before turns out to be the best. Two qualities that other engineers value the most are empathy and respect. If you have been an engineer yourself, it is easy to show empathy to other engineers. You just treat others the way you like to be treated, and don’t treat others the way you didn’t like to be treated. Engendering respect is something you have to learn to earn. Usually that comes from the decisions (technical or other) that you make, treatment of others and overall philosophy. If you manage to engender respect, you are on the way to inspire your team. That way, you can achieve the key to loyalty and high probability of success.</p><p><strong>Organization</strong></p><p>At a CTO conference (<a href="https://www.ctoconnection.com/">CTO Connection</a>) I first learned about “<a href="https://en.wikipedia.org/wiki/Conway%27s_law">Conway’s law</a>.” The law, named after a programmer named Melvin Conway says, “organizations which design systems … are constrained to produce designs which are copies of the communication structures of these organizations.” I have found out that this has been true, in other words, your team organization is reflected in your architecture. For a startup in its early days, I favored a flat organization where most (if not all) members are full stack and “interchangeable.” In terms of architecture, you end up with a balanced one, where there aren’t any extreme points of weakness.</p><p>Two organizational points (for a startup) that I liked to take care of from the beginning were on-boarding and code repository/code workflow. By on-boarding, I mean the blueprint or playbook by which a new member of the team gets started. For example, I once had the whole installation procedure (of the system) documented. A new engineer was able install the entire system on his first day without asking me any questions.</p><p>Code repositories and code workflows are things that have to be setup from the get-go. Although workflow procedures are not really needed until you have a team (more than one person).</p><p><strong>Curiosity</strong></p><p>Curiosity of the intellectual kind is for me the driver for innovation. Tech folks always seek newer and better ways to solve problems. As CTO, you have to be ahead of the curve. You have to be constantly reading about the latest tech, going to meetups and conferences. Also, you have to be challenging your team-mates to read up of something new and try it out. Tinkering. Breaking toys to see how they work are qualities that you had as a kid.</p><p>If you are not “out there” being curious about the world and what others have built or are building, you’ll most probably reinvent the wheel and waste a lot of time. This is time that you could be using on your core tasks or product.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j4071daIqOQeA5cewwmMTA.jpeg" /><figcaption>Jedi mind tricks are “soft skills”</figcaption></figure><p>Engineers are like big kids. They like StarWars, StarTrek, Pokemon, comics, puzzles, and board games. Playfulness is a key element for an engineer to thrive. If you let engineers “play” with new technology, you make it possible for innovations to flourish from the ground up. We applied this playfulness at quarterly internal <a href="https://ocw.mit.edu/courses/sloan-school-of-management/15-369-seminar-in-corporate-entrepreneurship-fall-2015/related-resources/MIT15_369F15_ThesisFinal.pdf">hackathons</a>. And speaking of board games, we used to play Pictionary in the evenings. One time we had a even clients and other entrepreneurs play with us. It was great for team bonding, collaboration and “get to know you better.”</p><p><strong>Imagination/Vision</strong></p><p>Disney engineers are very good at this, they can imagine a new world and later make it “real.” Not for nothing they are called “imagineers.” Imagining the “ideal world” is something that a CTO has to bring to the table. The imagination transforms into vision, which will inspire and drive your team.</p><p>I first “imagined” modeling interactions between users and products (for Open Sesame’s personalization engine) using directed graphs back in the early 2000s. During those days, Java and relational databases were the technologies <em>du jour, </em>thus I had to be constrained with the technology of the day. I could not model user interactions without making costly conversions. It wasn’t until 2012 that I got hold of a <a href="https://en.wikipedia.org/wiki/Graph_database">graph database</a>. This kind of database allowed me to model highly interconnected data like transportation networks. The result of this, was the work at <a href="https://www.wanderu.com/">Wanderu</a>. We used a graph database to route disconnected segments of buses and trains. From day one I had “imagined” (or conceptualized) the system and two years later it came to fruition, almost as I imagined it.</p><p>In summary, for a “C” level person like a startup CTO, communication, leadership and organization are obvious skills to have. For me, extroversion was a trait that did not have, but was crucial. And finally, in order for a CTO to “boldly go where no one has gone before”, one must definitely have curiosity and imagination/vision.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ca94a16f23a7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[7 Good Tech Practices to Follow (by Peruvian Startup Riqra)]]></title>
            <link>https://medium.com/@eddywongch/7-good-tech-practices-to-follow-by-peruvian-startup-riqra-d7af64cc15f8?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/d7af64cc15f8</guid>
            <category><![CDATA[peru]]></category>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Wed, 21 Mar 2018 14:26:28 GMT</pubDate>
            <atom:updated>2018-03-21T14:26:28.524Z</atom:updated>
            <content:encoded><![CDATA[<p>I sat down with Julio César Canares García, CTO and co-founder of startup Riqra, a platform of wholesale suppliers for SME (small and medium enterprises). Riqra is currently being accelerated at UTEC Ventures, one of the most well-know startup accelerators in Peru. I wanted to share with the startup community the many technical practices that he is following at his startup. Many of them, IMHO, are noteworthy for the Peruvian ecosystem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hXgabSAtJgOzjC7Z2c64qg.jpeg" /><figcaption>The Riqra github dashboard</figcaption></figure><p><strong>#1: English as a First Language</strong></p><p>I was pleasantly surprised that Riqra uses English throughout their development life-cycle. All is in English: variable names, file names, server names, documentation, comments, maintenance comments, bug reports, and even code reviews. Engineers who don’t know English won’t pass Riqra’s hiring filter, moreover, an English-speaking engineer will most definitely feel at home at Riqra.</p><p><strong>#2: Not afraid of Spending</strong></p><p>Many Peruvian startups suffer from “stinginess”, that is, they are not willing to spend hundreds of dollar per month on services (mainly SaaS that are available in another part of the world) that could, in the long run, save a lot of daily work, external consulting or headaches from badly-done custom work. The main concern on the part of (commercial side) startups is that when you compare the amount of funding that you are able to raise (hundreds of thousands of dollars for seed rounds) vs. the amount you are spending monthly (hundreds or thousands of dollars), it seems proportionally “too much.”</p><p>Julio chose the service <a href="http://heroku.com/">Heroku</a>, which is a cloud application platform that manages infrastructure (or PaaS — Platform as a Service). His bill runs in the hundreds of dollars, but Julio is happy with the productivity that he gets out of it. Instead of spending the time to create a custom-built system, he opted for this PaaS, which let him concentrate on building his product. In the early days, many startups waste too much time in building things rather than buying them and focusing on their core business. Many don’t make a good “buy vs. build” decision and try to tackle too many things at the same time.</p><p><strong>#3: Use the Best Technology for the Problem</strong></p><p>A pattern that many Peruvian startups show is their reluctance to step out of the comfort zone in terms of technology choice. Many startups have the chance to have a clean slate to start, but pick PHP (LAMP) as their technology stack. There is nothing wrong with LAMP, but the 90s was 20 years ago. Bucking this local trend, Julio had the courage to pick <a href="https://graphql.org/">GraphQL</a> instead of <a href="https://en.wikipedia.org/wiki/Representational_state_transfer">REST</a>, and <a href="https://nodejs.org/en/">nodejs</a> instead of PHP. This decision was made, with knowledge that there won’t be too much expertise (in GraphQL and nodejs) in the market to hire from. He will have to inspire and train them, as he has done so far. Julio couldn’t be happier with his choice of GraphQL, it abstracts the back-end so much, that he has “time to do other things”, he commented.</p><p><strong>#4: Having an Automated Code Workflow</strong></p><p>One of the most under-looked skills/knowledge in Peruvian startups is code organization. Often, engineers come out of school knowing how to make a project of one person, but not too many know how to setup project for a team of engineers and several environments. By setup, I mean organizing repository, branches, environments (development, testing, staging, production) and pipelines that go between them to deploy the right version the fastest way possible. In essence, having the whole process automated that works consistently without human error. At Riqra, they have set their development pipeline to release often, and they have sometimes done so, many times a day. They have organized their code around “feature flags”, meaning that they can release a feature, but don’t show the feature until the “flag” is on. Features can have their respective branches and server environments. In the Riqra workflow setup, one can create an isolated server with a new feature (for testing purposes) and load it with production data.</p><p><strong>#5: Understanding Agile, Really …</strong></p><p>Many startups, not just Peruvian ones, fall into the trap of following agile methodologies (like Scrum or Kanban) to the letter and forget that the most important goal is to come up with a product iteratively. A team cannot work iteratively if it cannot arrive to a “production ready” state often. This was well-understood at Riqra. They have avoided the fluff of the popular methodologies and adapted to a workflow that works for their tech team and their clients. A feature can be deployed to production literally in less than an hour. Their development workflow is proof of this understanding of moving in an agile way.</p><p><strong>#6: Developer Empowerment</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1eEuEOQpNiV7pzZT3erSZA.jpeg" /><figcaption>Riqra’s tech team celebrating their first signed client (at UTEC Ventures)</figcaption></figure><p>One aspect of development that I don’t see often in Peruvian startups is having an empowered tech team. Too often, I see the CEO calling all the shots or a CTO being a subordinate of the CEO, not a partner. In this kind of environment, developers are often under motivated and responsibility is misplaced. Julio understood this because of being a developer himself. He has set the Riqra development environment so that each developer is responsible for not only writing, but also deploying, monitoring, evaluating, rolling back and fixing their own features in production. They become truly “owners” of their features. For this purpose, he has set tools like <a href="https://www.pingdom.com/"><em>Pingdom</em></a> and <a href="https://www.fullstory.com/"><em>Fullstory</em></a><em> </em>to monitor performance and usability respectively. With FullStory, he told me that developers have watched how their feature was used by their clients, like a movie, even at wee hours at night. “This brings a sense of pride, when you are able to see your work being used in Production”, Julio asserts. In this way, by giving ownership and transparency, developers have a direct line to clients, and therefore feel more responsible for their outcomes.</p><p><strong>#7: Technical Leadership</strong></p><p>One manifestation of what I’ve mentioned above (lack of empowerment of the tech team) is that technical leadership is lacking in many Peruvian startups. The reasons are perverted, for once, many CTOs lack the technical leadership from the get go or just are not given the chance to lead, for fear that they might screw up or simply because they don’t have an assertive enough personality (most of us techies suffer from shyness). This is not what you find at Riqra. In order to be successful in an organization of techies, one must have empathy and understand how developers think. “Developers don’t like to be lead by somebody that they don’t respect”, told me Julio. At Riqra, Julio has set an environment and the processes that are transparent and give accountability to developers, and it’s a place where the work is interesting and cutting-edge, so that you stay motivated and challenged.</p><p>Among Peruvian startups, Julio has emerged as clear leader of his kind, because he has done things that are not commonly done in this part of the world. “I’ve learned mostly on the Internet … necessity is my driver”, says Julio. A technology leader is not only a master of all things technical, but also a master of soft kills like leadership, empowerment, vision, inspiration, empathy, transparency, accountability, motivation, responsibility. “The moment you start wearing the business hat, that’s when you’ve graduated to being a technology leader”, remarked Julio, in closing.</p><p>With all these practices, Riqra, with Julio at the technical helm, is on its way using world class development standards, mostly learned on the Internet. And one day, they might surprise us (the Peruvian ecosystem and the World) in many other ways.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d7af64cc15f8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Observations of the Peruvian Startup Ecosystem - Part 2: Startup Communities and the Emergence of…]]></title>
            <link>https://medium.com/@eddywongch/observations-of-the-peruvian-startup-ecosystem-part-2-startup-communities-and-the-emergence-of-e8387fc03be8?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/e8387fc03be8</guid>
            <category><![CDATA[peru]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[community]]></category>
            <category><![CDATA[network]]></category>
            <category><![CDATA[startup]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Fri, 04 Aug 2017 11:14:19 GMT</pubDate>
            <atom:updated>2023-06-20T01:17:20.018Z</atom:updated>
            <content:encoded><![CDATA[<h3>Observations of the Peruvian Startup Ecosystem - Part 2: Startup Communities and the Emergence of Arequipa</h3><p>In one of my visits to Peru in 2015, Rob McMillen, the then CEO of Alta-El Dorado Emprendimiento, a VC company in Lima, shared with me something surprising about the emergence of Computer Science in Peru. “You have to meet the Cuadros (Ernesto and Alex) brothers in Arequipa”, he stated. I replied incredulously, “Arequipa, really?” “They created the first Computer Science faculty in Perú”, he followed. I was amazed and intrigued, because everyone assumes that things like that happen in Lima, the capital. I was introduced by email and immediately Ernesto and Alex Cuadros extended an invitation to visit the Universidad Católica San Pablo. The actual meeting did not happen until several months later. I had encountered an unexpected “gem” in the Peruvian Ecosystem, a gem that makes Arequipa a prime candidate to become a hub for computer science, technology and innovation (and therefore, tech startups).</p><p>In this post I will describe the ingredients that I have found that makes Arequipa a prime candidate for a thriving Startup Community, with plenty of observations and recommendations.</p><h4>Origins: The Ernesto Cuadros Story</h4><p>In 1995, Ernesto Cuadros arrives to the University of São Paulo, after a multi-day ride by bus, with $400 in his pocket and a heavy backpack. The goal: to study Computer Science and escape the backwardness of Perú. About 10 years later, in 2004, after having lived in Brazil, Germany and USA, he launches a brand new field that he studied: Computer Science. At the Universidad Católica San Pablo, he becomes the director of the first Computer Science program in Perú modeled under the guidelines of ACM (Association for Computer Machinery). The Bachelor’s program lasts 5 years, with (unusual) requirement of a thesis sustentation, in most cases in English, and sometimes on Portuguese. In addition, he creates an environment of camaraderie where professors are called by their first names and a work ethic where everyone is self-reliant and everyone helps everyone else.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jhDmPxxWYuGEUm9vOBh6PA.jpeg" /><figcaption>School of Computer Science at San Pablo</figcaption></figure><p>Ernesto not only founds the first Computer Science program in Peru, but also establishes a “bridge” to Brazil. While he was a student there, he encouraged his brother, Alex, and his friends to come to Brazil to obtain PhDs with scholarships granted by Brazilian institutions. This mini-migration ends up conforming the bulk of the faculty at San Pablo which counts with 90% of the PhDs in Computer Science in Perú. This also has the healthy effect of opening doors of graduates of San Pablo, at least tens of its graduates go on to pursue a Master’s or a PhD in Brazil every year, because of the Brazilian connection.</p><h4>Startup Communities relies on Networks (aka Graphs)</h4><p>Brad Feld, a well known entrepreneur, venture capitalist, cofounder of Techstars affirms that entrepreneurial ecosystems require networks as opposed to hierarchies to thrive. In other words, the formal and informal connections that are formed between entrepreneurs and mentors is critical for entrepreneurial education. Ernesto Cuadros had willingly or unwillingly made two connections. First, Peruvian students to Computer Science and secondly, to opportunities for advanced degrees in Computer Science in Brazil. In network-speak, he became a “bridge” or enabler that stands “between” a student and an opportunity in Brazil.</p><p>But, what is a Network? Also known as a “graph”, is a mathematical structure used to model pairwise relations between objects. The academic field that studies graphs is called graph theory, and a superset of this field is called Network Science. It incorporates theory from graph theory, <a href="https://en.wikipedia.org/wiki/Statistical_mechanics">statistical mechanics</a> from physics, <a href="https://en.wikipedia.org/wiki/Data_mining">data mining</a> and <a href="https://en.wikipedia.org/wiki/Information_visualization">information visualization</a> from computer science, <a href="https://en.wikipedia.org/wiki/Inferential_statistics">inferential modeling</a> from statistics, and <a href="https://en.wikipedia.org/wiki/Social_structure">social structure</a> from sociology.</p><p>That’s right, there is a whole “science” behind Networks, and one can even obtain a PhD in Network Science. But why is this important for somebody involved in startup ecosystems, for example? If you study Network Science, you will learn that networks, specially of the social kind — the one in a Startup Community — grow in particular way, it’s called “preferential attachment”, ie. in a social network the nodes with lots of connections will continue to grow rapidly, whereas new nodes will be slow to form connections. Translation: If Ernesto had not created an egalitarian environment at San Pablo, very few students would know or even would have the chance to pursue advanced degrees in Brazil. He is the node with the high connection count, the “bridge”.</p><p>Other examples of how Network Science can help understand Startup Communities is by studying “knowledge diffusion”, how knowledge disseminates within a network and “community formation”, how a community automatically emerges after certain connections are achieved. But those are deeper subjects for another article.</p><p>Knowing the basic theory about networks, an organizer of a startup entity like an incubator or accelerator, can better plan for the type of events where newcomers of a network (the new cohort) can bond to people in the network that already have lots of existing connections. Similarly, “communities”, including informal ones, form themselves, if the people involved have a common purpose or something to share. In a network, a communication between two nodes is the fastest, whereas in a hierarchy, it is slow, difficult, if not impossible going bottom-up.</p><h4>Exporting the Gold</h4><p>The results of the Computer Science faculty at San Pablo showed its fruits years later. Many of the students managed to obtain internships at world-class companies like Google. And several years later, several of them even ended up being hired for full-time positions, at Google. It was such a win for the school and the country that it appeared on several news <a href="http://www.fsnoticias.org/noticias/dos-egresados-la-universidad-catolica-san-pablo-trabajan-google-14981">outlets</a>.</p><p>This was a moment of validation for Ernesto Cuadros as the head of the CS faculty at San Pablo. The hardships trekking all the way to Brazil, wanting to study Computer Science so bad, and later to return with his friends to form a faculty in Peru paid off, not without headwinds and the naysayers.</p><p>However, this short term victory for the students at San Pablo can be seen as a loss, rather than a win, strategically speaking. Peruvians usually chant victory when we increase the production of gold. In a similar vein, we chant when we “export” the best students to Google. In reality, it is a loss for the country since it means a brain drain of the best. In addition, it is not a realistic goal for the rest of the students to aim for Google, because not all of them can turn up there. Not all of them will be capable or willing to pursue a PhD in Brazil either. The ones that make it to Google or Brazil are probably in the single digits. There has to be other alternatives for the rest of the graduates.</p><h4>The Chicken and the Egg</h4><p>The alternative for the graduates of San Pablo is definitely not the traditional Peruvian IT industry, with its high reliance on Windows, Oracle, and proprietary systems. The typical profile of a CS graduate at San Pablo is Unix, Python, C++/Java, solid CS foundations, open source and a couple of internships under his/her belt. Many graduates “pass thru” a local financial startup, InkaLabs, which is becoming the facto professional training ground in Arequipa. Most CS students at San Pablo either intern there in the Summer or work part time. Although more startups are appearing locally, there isn’t enough opportunities to absorb the local supply of CS graduates. Many end up working at startups in Lima (San Pablo students are very sought after by Lima startups) or end up working remotely for foreign companies, mostly from the US and Uruguay. In fact, there is a sizable community of “remoters”, probably more than 30 in Arequipa alone. The most notable group is PeruCoders. The alternative of remoting is so appealing for Arequipeños because they end up using open source technologies, many offer flex-time, the pay is at least 30% above market and it provides a ticket to working abroad, another path to “exporting gold”, unfortunately.</p><p>Why aren’t there more startups? In one word, economics. If you are a recent graduate at San Pablo, and you are seeking a better quality of life, your top choices are 1) Google 2) Advanced degree in Brazil (if you get a scholarship) 3) Work remotely for a foreign company 4) Lima startups 4) Arequipa startups. Being part of a startup is more attractive during the years of school than after graduation, since the incentives and attractiveness are few. In terms of money, #3 is the most appealing. However, this creates a community of “islands”, that have no connections to others. In others words, it does not create a healthy community.</p><p>On the industry side, things are moving toward the right direction, but slowly. The ideal situation is for existing industry to start waking up and peek into new ways of doing things with startups. One laudable effort has been spearheaded by Jean Louis Gelot, a former French exec, President of the IT Committee of the Chamber of Commerce in Arequipa. He recently organized the first ever <a href="https://web.facebook.com/events/230282854150993/?acontext=%7B%22ref%22%3A%223%22%2C%22ref_newsfeed_story_type%22%3A%22regular%22%2C%22feed_story_type%22%3A%22263%22%2C%22action_history%22%3A%22null%22%7D">Peru Digital</a> conference in Arequipa. The event was well concurred by industry as well as students and entrepreneurs.</p><p>Another case of laudable effort that should be recognized was by a startup called <a href="https://emesh.la/">eMesh</a>, and event management service. They organized a “lateral” event where they shared their experiences at StartupPeru with others in the ecosystem. The event was called <a href="https://web.facebook.com/photo.php?fbid=1572272909452259&amp;set=a.431380503541511.104626.100000086967165&amp;type=3&amp;theater">eMesh Conecta</a>. I remember that when they attended the meetup #MeetupDeLosTechies, a meetup that I organized in Arequipa, I challenged their team to “return the favor” by sharing their own experience. They did, this is an example to follow by other startups. Give before you Get.</p><p>Unfortunately, angel investment as such has not shown much activity. VC investment is practically inexistent, as far a I know. This part of equation still has ways to go.</p><p><strong>Conclusions and Recommendations</strong></p><p>Arequipa enjoys an enviable position of being the producer of perhaps the best talent in Computer Science in Peru. This “resource” needs to be nurtured for it to grow and be productive for the country. Government has a big role in enabling and giving incentives for industry, angel, and VCs to invest or co-invest. Following are some conclusions and recommendations:</p><ul><li>The environment that Ernesto Cuadros created and the work ethic where everyone helps everyone else are key for the DNA of a startup ecosystem. Give before you Get. This is a mentality that is new for Peruvians. The traditional Peruvian mentality is Take, not Give.</li><li>Open source software, used at San Pablo, is the right path for technology in Peru. Proprietary systems do not belong in Peruvian universities or even the government. They are an enabler of possible corruption. Open source and open data forces transparency and possible temptation of self-dealing with software vendors. Learn from the Odrebrecht scandal, a multimillion corruption case that embroiled even former Presidents. Open source and open data are the right path for government, learn from the <a href="https://www.usds.gov/">US Digital Service</a>.</li><li>Stop the “export of gold.” The Peruvian government has to create incentives for talent to stay, or to come back after a number of years. If students go abroad and come back they create an automatic “bridge” to the country they went. Ernesto and Brazil is an example. In India, many of the startups have links to VCs in Silicon Valley, where many Indian entrepreneurs and technologists managed to thrive. The mantra should be: “Export (computer) monitors, not the raw gold that comes inside.”</li><li>There has to be something else other than “aiming for Google.” It is not realistic that this is what all graduates could reach. There has be more opportunities that give graduates a decent quality of life. IT salaries are a fraction of world markets and people take advantage of it.</li><li>One missing link is industry. Their participation is lukewarm. The easiest way for (big) industry to participate in the ecosystem is to donate their event rooms for community events after hours.</li><li>CS graduates need to mix with business hustlers (business majors) to create more startups. Angels need to invest or co-invest at seed levels, and not demand too much equity, so that founders feel that they are getting a fair deal.</li><li>A soft-landing center is needed to provide more opportunities to San Pablo graduates and become a training ground of software best practices. Government should bring foreign startups and experts to visit and stay for a period of time. A co-working space next to affordable lodging could be a great idea. This could benefit the rest of the country.</li><li>There are too few VCs and too few angels that invest. Also too few mentors, too few judges, that understand what a tech startup is. They haven’t done one themselves. The chicken and the egg situation continues.</li></ul><h4>References</h4><ul><li><a href="http://sodalicio.org/noticias/un-peruano-que-cambia-la-historia/">http://sodalicio.org/noticias/un-peruano-que-cambia-la-historia/</a></li><li><a href="http://ecuadros.blogspot.com/2014/11/el-secreto-de-cs-ucsp.html">http://ecuadros.blogspot.com/2014/11/el-secreto-de-cs-ucsp.html</a></li><li><a href="http://jornal.usp.br/universidade/ele-lancou-novas-bases-para-as-ciencias-de-computacao-no-peru/">http://jornal.usp.br/universidade/ele-lancou-novas-bases-para-as-ciencias-de-computacao-no-peru/</a></li><li>Startup Communities by Brad Feld, <a href="https://www.amazon.com/Startup-Communities-Building-Entrepreneurial-Ecosystem-ebook/dp/B008UV826U/ref=sr_1_4?ie=UTF8&amp;qid=1499480955&amp;sr=8-4&amp;keywords=brad+feld">https://www.amazon.com/Startup-Communities-Building-Entrepreneurial-Ecosystem-ebook/dp/B008UV826U/ref=sr_1_4?ie=UTF8&amp;qid=1499480955&amp;sr=8-4&amp;keywords=brad+feld</a></li><li>UCSP at Google: <a href="http://www.fsnoticias.org/noticias/dos-egresados-la-universidad-catolica-san-pablo-trabajan-google-14981">http://www.fsnoticias.org/noticias/dos-egresados-la-universidad-catolica-san-pablo-trabajan-google-14981</a></li><li><a href="https://en.wikipedia.org/wiki/Network_science">https://en.wikipedia.org/wiki/Network_science</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e8387fc03be8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Observations of the Peruvian Startup Ecosystem]]></title>
            <link>https://medium.com/@eddywongch/observations-of-the-peruvian-startup-ecosystem-bdfba12b46f2?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/bdfba12b46f2</guid>
            <category><![CDATA[entrepreneurship]]></category>
            <category><![CDATA[latin-america]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[peru]]></category>
            <category><![CDATA[tech]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Fri, 16 Dec 2016 04:13:40 GMT</pubDate>
            <atom:updated>2023-04-19T16:28:24.323Z</atom:updated>
            <content:encoded><![CDATA[<p>Last month I had the chance spend several weeks in Peru (Lima and Arequipa) as speaker and guest at universities, incubators, community events and startup-related organizations. I witnessed first-hand how the tech startup ecosystem is evolving. I saw very good signs and enthusiasm focused in the right direction. These are my observations with a <strong>unique</strong> perspective … I was in born Lima, Peru.</p><p>Although I am a product of Peruvian private elementary and secondary education, I never attended a Peruvian university, nor held a job in Peru. I left Peru right after high school and became schooled in Computer Science and Math in the US. Afterwards, I worked in the US software industry (startups, small, medium, large corps) for more than 20 years. One could say that I had a Peruvian childhood and a <em>gringo</em> adulthood. And although I admit spending more time as a <em>gringo,</em> I can still detect the little nuances and idiosyncrasies in the Peruvian way of thinking. Here’s my personal assessment of the ecosystem with one American professional eye and one Peruvian “psycho-analytical” eye.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hIUP0O6u9QGzfkL9GlaJkw.jpeg" /><figcaption>At StartupGrind Lima</figcaption></figure><p><strong>No Technical Cofounders</strong></p><p>After one of the events I attended, I stopped by a fairly publicized and recently funded startup. I engaged in conversation with one of the team members of this startup. “So tell me about your tech stack”, I asked casually. “It’s PHP and MySQL, but it’s not the best”, he replied. “Why do you say that?”, I countered suspiciously, not expecting that answer. “It was chosen by a third party”, he answered. It turns out the startup began without a technical cofounder and the CEO outsourced the solution to a third party.</p><p>Apparently this is more the rule, than the exception. For some reason (which I’ll to elaborate later) technical people tend not to reach the leadership sphere in Peru’s tech ecosystem. Most startups get launched by their would-be CEOs. But, they are trying to build <em>tech startups </em>right? Would you ever open a restaurant without a chef?? Would ever outsource your chef?</p><p>The startup system itself promotes this because of the emphasis on pitching competitions. Who raises to the top in a pitching competition? The one with the most concise and eloquent pitch, and charismatic personality, definitely not traits of a tech person. This emphasis on pitching occurs in US ecosystems as well; however, in the US there are tech communities that are peer-driven that recognize your work for substance — your actual work (Github, tech conferences), not just presentation. This peer-driven community does not exist, or is very incipient at best, in Peru.</p><p><strong>The Imbalance between Biz and Tech People</strong></p><p>Growing up I remember that the most popular careers were “Administración de Empresas” (business) and “Ingeniería Industrial” (industrial engineering). In both you were groomed to be the next “empresario” (enterprise boss) and if you went to the colleges with the most reputation, like Pontificia Universidad Católica del Peru, Universidad de Lima or Universidad del Pacífico, you were for sure on track to be next the CEO of a Peruvian enterprise. I asked to myself then, if everybody aimed to be bosses, who is going to be left to be a boss of? Or if everybody focused on industrial engineering, what industry would they work on? Because there wasn’t a lot of industry back then. Perhaps a whole generation was trained to be bosses, not builders or makers.</p><p>At the same time that I was leaving for the US, a number of academies or institutes appeared in the Lima market. These offered two-year degrees in “computación” or “informática”, mostly akin to a technical degree in MIS (management information systems). This is one of the primary reasons I left Peru, because the lack programs in computer science (CS), the study theory, methods and techniques that make computers work. Fortunately I had traveled to the US before, and I knew of the existence of CS.</p><p>You can see how these two groups interact in the formation of startups. For sure, the union is not one of equals. Why is this? I asked myself. I have a theory. Back in the 80s, during my college years in the US, I was an intern at GTE (now Verizon), a major US phone company. I remember that all my bosses seemed to have backgrounds in business. By the 90s, more and more bosses tended to be have technical background. What happened? The shift from project-management driven, Gantt charts and waterfall methodology to a more peer-driven, agile methodology was transforming how tech projects got planned and delivered. It also affected how organizations were structured around tech teams, giving more responsibility and leadership to tech people. This peer-to-peer methodology or approach has even expanded beyond single corporations or institutions, to collaborations across teams and across countries. Moreover, the open source movement and the commercial success of open-source-driven software has accelerated this way of collaboration.</p><p>In Peru, the shift to agile and open source has not happened. In the US, during this transformation, the leadership (power and responsibility) shifted towards tech people. This has not occurred in the Peruvian industry, that is why tech people are still second class citizens, still taking orders from business people and therefore, not getting the chance to rise to leadership spheres.</p><p><strong>The Elephant in the Room</strong></p><p>One of the things that was appalling to see is how big software package vendors are “embedded” in the industry, government and even higher education. The strategy of these software vendors is to create a dependency early on, at your university for example, on these software packages. The purpose is to perpetuate the use of these packages later on in your career, thus preventing you from creating anything on your own. On the government side, it is easy for government officials to shift the responsibility to third parties and have the option to “blame” this third party when projects don’t work out. In education, the big software vendors “donate” a computer center and free licenses to university and schools. In my opinion, an institution of higher education should be a place of universal thinking, where a particular commercial solution has no place as a requirement, perhaps as subject or method of study, among others. This is especially important in computer science or computer related fields where creating something from scratch is so easy. You are just killing innovative instincts.</p><p>What does this dependance on big software packages create? It creates a “software” community composed of mainly implementation, maintenance and support engineers. A community that will not create anything from scratch, because it doesn’t know how to. This flaw in the community is insidious and permeates itself into the startup industry and into higher education. Would you want somebody that does not know how to create things from scratch in a startup? This is a problem that is a back-breaker in a startup ecosystem. If you are a startup and you are in need of experienced technical talent (post Series A for example), the Peruvian (corporate) industry will only offer you less than ideal candidates, well versed in software packages.</p><p>This dependance on big software packages rears its head in higher education as well. In this industry, since you only need implementors, maintainers and support people to fill the corporate ranks, you favor programs that fulfill that need: Management information systems (MIS) or “ingeniería de sistemas” as known locally. Most computer-related university programs are oriented to churn engineers that “use” software, not to build software. Only in the last ten years, have computer science programs been created, and of all places, the first one happened to be in Arequipa, the second largest city of Peru, as opposed to the capital, Lima, where “everything” important usually happens. This is an example of successful out-of-the-box thinking, not centralized group-think.</p><p><strong>There are plenty of Problems, where are the Solutions?</strong></p><p>When you see a the list of startups at various incubators you mostly see “tropicalization” (local version) of foreign ideas (ecommerce, social networks) and lots on “me-too” or lifestyle-oriented startups. There is a glimmer of hope as you are starting to see some startups oriented toward problems that are focused on primarily Peruvian or Latam problems, like non-traditional banking, and gastronomy. However, a lot of roadblocks in the Peruvian way of thinking conspire against better ideas.</p><p>Growing up in Peru, I saw a culture of conformity and “obedience” to “authority” and the status quo. Also, the believe that anything “good” comes from a foreign country. This must be a reflection of a history influenced by the fact of being conquered by Spaniards for more than 200 years, a regimental Catholic church that influenced our customs, and more recently, decades of military governments, and restriction of imported ideas and goods. Lest we forget, that the liberation from the Spaniards came from movements in Argentina and Venezuela, <strong>not</strong> from within. Non-conformity or creativity <strong>(</strong>of ideas) is not only not encouraged in Peruvian culture, but curtailed from childhood. You see that in the education system, with the emphasis on recitation, memorization and strict imitation. You almost see an <strong>insecurity</strong> in the Peruvian psyche, the thinking that “it can’t be done and the solution must come from overseas.” This is completely opposite from the US philosophy of questioning the status quo, or thinking “is there a better way?” This is something embedded in the DNA of the US, the existence of the United States was based on that thinking, that men and women are created equal, that there is no king, that church must be separated from state and overall, that their own citizens are responsible for forging their own future.</p><p>Another aspect of Peruvian way of thinking is the emphasis on image, rather than substance. People like to be associated with “famous” people in photos. Or, they go to universities or workshops to earn a diploma (“el carton”) and think that the piece of paper is their most important achievement. Or, get into debt to buy car and clothes, to appear “more” before your colleagues and friends. Or, think that somebody dressed in a suit and tie is more credible. This way of thinking conspires against the way new companies should be created. You can see this manifest in startups: the pitch, website and message will be extremely polished, ie. image, but not the tech stack or business model behind it. This is a reflection of image-first thinking.</p><p>Entrepreneurship and ingenuity are not foreign ideas in Peru. You just need to look at the thousands of street vendors around Lima, or anywhere in Peru. The street vendors have figured out a way to make money, sell something at higher price than cost, and know how to acquire customers by pounding the streets. Ingenuity, you could see in many places, even during my childhood times. The “Avenida Azángaro” was a well-known street in downtown Lima, for forging any sort of documents to an uncanny reproduction. They even did almost authentic university diplomas. Ingenuity and entrepreneurship in Peru have been miss-directed for decades, if not centuries, and have not had the chance to be combined with the right know-how.</p><p><strong>Moving forward, some ideas to make the Ecosystem better</strong></p><p>If I had a magic wand, here are some ideas on how to improve the startup ecosystem in Peru:</p><ul><li>Community — Encourage and build grass root efforts, peer-driven communities, not mandate top-down. Peer communities by nature are self-correcting, because there isn’t a hierarchy that becomes ossified, and new members are added at an equal level.</li><li>Meritocracy — Reward merit, substance over form. Don’t just do pitching competitions, but also hackathons and <a href="https://www.kaggle.com/">Kaggle</a>-style problem solving competitions. Don’t just look at diplomas, resumes, LinkedIn, CV, but also at Github and projects.</li><li>Open source and Agile — Government institutions should adopt (only) open source technologies and demand agile delivery of projects. Big software vendors cannot cry fowl, because they’ve adopted open source already in the US. Just don’t keep buying 20-yr old solutions.</li><li>Education — Emphasize computer science education from early age, not just MIS, that is, teach everybody how to program and think analytically, not only use software packages. Focus on creating more computer science departments in universities, not just MIS, focus on creating not just using.</li><li>Build Exchanges of Ideas—Look for ideas from outside first, emulate them and nurture them so that eventually you can create your own. Invite people with those ideas. Repatriate those who once fled the country.</li><li>Imagination and Creativity — Go (or follow) to Disney, volunteer as an Imagineer and see how they do things. Mix art and science, like how gastronomy mixes chemistry and presentation, or how tourism can mix archaeology with data science.</li><li>Learn from the street vendor or the informal economy — Make startup entrepreneurs work as a street vendor as a way to learn entrepreneurship. This is akin to “selling lemonade”, the usual example to illustrate childhood entrepreneurship.</li><li>Learn from the Gastronomy industry — You see a myriad of new dishes and restaurants in Lima and all over Peru. But they are not just following recipes, correct? They are creating new recipes. So what do they teach in culinary school? They teach principles of how different ingredients are prepared and combined to maximize flavors and textures. Do we have an inferiority complex? No. We proudly claim it’s one of the best cuisines in the world and we’ve shown it. Now …</li></ul><p>Similarly, a computer science department in Peru should teach you the principles and concepts behind how computers and programs work. So that, the new generation of Peruvians are able to create the new “dishes” that solve world problems that we could export to the rest of the world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bdfba12b46f2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Apple vs. Google, a War of Content]]></title>
            <link>https://medium.com/@eddywongch/apple-vs-google-a-war-of-content-ded23dfe677f?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/ded23dfe677f</guid>
            <category><![CDATA[ad-blocking]]></category>
            <category><![CDATA[advertising]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Fri, 29 Jan 2016 15:35:33 GMT</pubDate>
            <atom:updated>2016-01-29T15:35:33.837Z</atom:updated>
            <content:encoded><![CDATA[<p>Apple vs. Google, a War of Content</p><p>[Originally published: May 26, 2010]</p><p>Most technology pundits have said that the rivalry between Apple and Google started when Google got into the cellphone market in direct competition with Apple’s iPhone offering. The reality is that the rivalry is deeper than that, the rivalry is about open content vs. proprietary content.</p><p>Google’s success came about on the web. Millions of websites sprung up during the Dot-com era to the point that nowadays every self-respecting business, organization or club has a website. There was a mad rush to be on the web, most of the time, offering the content for free. Web users got accustomed to free content. Content on these websites were mostly written in HTML, a format that allows you to peek into the text itself and even the structure and formatting. That is, even if you write “exclusive content”, anybody can cut and paste or plagiarize your copy. Because the text of this content can be easily read by humans or machines, Google had a relative easy job in creating a search engine for the entire web. By “relative easy”, I mean, Google did not have to create the content, convert it or spend a penny obtaining the content. The content was just there to be read. Google then, to its credit, developed something that most of its search engine competitors have not done at the time, a business model. Google was the first to develop “paid search”, ie. charging money for a preferred position in a search result. This revolutionized the search market.</p><p>Years later, Google came up with “AdWords”, an advertising mechanism that allowed content providers display ads created by Google and get paid. These ads were auctioned to advertising businesses based on keyword selection and price. AdWords became the model for advertising on the web, filling the coffers of Google. With paid-search and AdWords, Google had essentially “killed” the business model of every newspaper and magazine on the planet. In essence, as a result of leveraging free content.</p><p>In 2001, when Napster and music file sharing was all the rage, Apple came up with iTunes, a proprietary marketplace that sells a song for 99 cents. Apple came up with a business model where there was none. This parallels Google’s coming up with a business model for search. With the introduction of the iPhone in 2007, iTunes not only became a successful music marketplace, but a marketplace for iPhone applications or “Apps”. iPhone apps allowed content creators create their application experience with their own graphics, copy and interactions. In other words, iPhone apps have created a fence to Google, no longer can Google sneak in and parse my content or display ads on a corner. The public Internet had now seen the creation of successful gated community where everybody wants to go, but that is controlled by Apple, not Google.</p><p>So the battle continues. In retrospect, most companies (including companies that sell content for a living) rushed to gates during the Dot-com era to offer their wares for free and forgot to charge for it. As a result, they have gone bankrupt and money has gone to Google. Google is now more of a media company than a search company, all their money comes from advertising. Apple apps have become a “second chance” for companies to think about about their business model before giving up their content for free.</p><p>Come to think of it, the aim of HTML was not to make money, it was invented by researchers at CERN to open up every piece of content. Content providers have just committed suicide when they went to the web before thinking about their business models and giving up all their money, first to Google … and then to Apple.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ded23dfe677f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Will BigData cause BigLaziness?]]></title>
            <link>https://medium.com/@eddywongch/will-bigdata-cause-biglaziness-4fabc865a643?source=rss-90e3eb1bdb60------2</link>
            <guid isPermaLink="false">https://medium.com/p/4fabc865a643</guid>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[coding]]></category>
            <dc:creator><![CDATA[Eddy Wong]]></dc:creator>
            <pubDate>Tue, 19 Jan 2016 18:19:40 GMT</pubDate>
            <atom:updated>2016-01-19T18:19:40.438Z</atom:updated>
            <content:encoded><![CDATA[<p>[Originally published: May 12, 2012]</p><p>Remember the days when Moore’s Law was telling us that CPU speed and RAM capacity was doubling every 18 months or so? What if we tell programmers that your compute power (size of your cluster and throughput) can be doubled every year and that the tools for doing big computations are so easy to use that anybody will able to submit such jobs? Welcome to Compute Laziness.</p><p>When RAM became cheaper and cheaper, programming techniques became “lazier”, ie. memory efficiency went out the window. Programmers instantiated huge objects on the heap, read entire XML files into memory for parsing, created data structures that worked entirely in memory. Gone were the days of packing every single useful bits, like using an unused area of your bytes for some other purpose. If you think about it, a byte can represent 256 (2⁸) different values, yet the English alphabet only has 26 letters. Granted there are other punctuation symbols, but in general any piece of text is, I would say, 90% made of the alphabet. Do you realize how wasteful we have become in the way we process data and the way store data?</p><p>There is a connection between economics and user behavior that is prevalent not only in computer science, but also in any human activity. When gas was $1/gallon in the late 90’s what did we do? We bought SUVs, built more highways, created more drive-thrus, created more suburbs that were far from city centers, etc. Now that we are back to $2–4/gallon we are suffering from those bad decisions in infrastructure. That existing infrastructure was built assuming that gas was cheap. What do you do to go from point A to point B at a cost equivalent to $1/gallon? You use more efficient cars, use more public transportation, and pack more living space near public transportation hubs. However, changing infrastructure is not an overnight thing, it takes years to transition, if at all.</p><p>Similarly when you are building your computing infrastructure, you might assume that a certain resource is cheap and that it follows Moore’s Law. Look at what happen in 2012, when floods in Thailand cause hard-disk scarcity. By the same token, mobile caught people in the following way, memory and CPU efficiency was back because you wanted to preserve battery life. So if you created and algorithm that was CPU intensive during the days of plenty CPU and RAM on the desktop, you probably couldn’t use it in your mobile application. In fact, that was one of the main complaints of Steve Jobs about Flash, it was CPU intensive and it would kill your battery life. Come to think about it, I see that once in a while on my desktop browser on pages with Flash, ie. “unresponsive Javascript” error.</p><p>So coming back to the question of BigData. What’s going to happen if we make compute power for big tasks like going thru log files and scouring user history very cheap and easy to use? People will undoubtedly start doing things using the brute-force way like using O(n^2) or O(n^3) or worse algorithms. So don’t get caught when BigData goes back to the equivalent of $4/gallon. The Earth has limited resources and Moore’s Law cannot go on for ever. The same way we’ve invested in solar and wind, we should invest in sound and efficient algorithms, not brute-force ones.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4fabc865a643" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>