tati
/
deiv-resume


			
				
					
						
						
							123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221
							<!DOCTYPE html>
<html>
<head>
  <title>David Ventura - resume</title>
  <link rel="stylesheet" href="css/style.css">
  <link href="https://fonts.googleapis.com/css?family=Lato|Open+Sans|Montserrat:400,400i,600,700,800&display=swap" rel="stylesheet">
</head>
<body>
  <header>
    <div class="resume-title">
      <h1>David Ventura</h1>
      <h2>
        SysAdmin, Devops & Dev
        <small>Jack of all trades</small>
      </h2>
    </div>
    <div class="personal-info">
      <p>
        Argentinian living in The Netherlands <br/>
        <a href="mailto:davidventura27@gmail.com">davidventura27@gmail.com</a> <br/>
        <a href="https://github.com/davidventura">github.com/DavidVentura</a> <br/>
        <a href="https://gitlab.com/davidventura">gitlab.com/DavidVentura</a> <br/>
        <a href="https://blog.davidventura.com.ar">blog</a> <br/>
      </p>
    </div>
  </header>

  <section class="experience-category">
    <h2 class="experience-title">
      Work Projects
    </h2>
    <div class="work-experience">
      <h3 class="company-name">
        <small>2018 - current</small>
        Optiver
        <small class="position">Sysadmin - Tooling developer</small>
      </h3>
      <p>
        Enforced operational commitments from development teams to bring down application errors to monitorable state.
        <ul>
          <li>
            Overhaul compliance related backup system (300TB+)
          </li>
          <li>
            Deployment system for production services (11k services, 860 servers)
          </li>
          <li>
            Linux-on-windows initiative and platform deployment to standardize the production API across the existing services
          </li>
          <li>
            Log and event analysis to quickly detect production related incidents
          </li>
          <li>
            Redesigned and implemented market data permissioning system
          </li>
          <li>
            (Currently) Redesigning server deployment
          </li>
          <li>
            Mentored colleagues, healping them transition from an old-school sysadmin position to a tooling development position
          </li>
        </ul>
      </p>
    </div>
    <div class="work-experience">
      <h3 class="company-name">
        <small>2014 - 2018</small>
        Especificos Buenos Aires
        <small class="position">Sysadmin - Tooling - Backend dev - Frontend dev - Embedded - Networking</small>
      </h3>
      <div class="company-project">
        <h4>Live-streaming Platform</h4>
        <p>
          The objective of this project was to create a complete pipeline for live video streaming, all the way from the physical cameras in the classroom to the software used by the alumni. A few interested points covered by the project:
          <ul>
            <li>
              Embedded linux systems for the video cameras (raspberry pi), audio mixer and slide system
            </li>
            <li>
              Live audio mixing of remote and local speakers
            </li>
            <li>
              Time syncing across video sources to prevent frame drift, using PTP
            </li>
            <li>
              Slide software that allows presenter to draw on slides and alumni to ask questions to the presenter
            </li>
          </ul>
        </p>
      </div>
      <div class="company-project">
        <h4>Online training platform</h4>
        <p>The objective of this project was to create a ‘forum’-like website to post the videos with the corresponding text material, allowing the students to communicate with each other.</p>
      </div>
      <div class="company-project">
        <h4>
          Billing System
        </h4>
        <p>This project manages most of the finances of the company and some extras:</p>
        <ul>
          <li>Billing, receipts (AFIP), shipping labels, e-commerce, stock and supplies</li>
          <li>Document scanning, printint, security cameras, certificate generator and monthly reports</li>
        </ul>
      </div>
    </div>
  
    <div class="work-experience">
      <h3 class="company-name">
        <small>2014-2018</small>
        Freelancing
      </h3>
      <div class="company-project">
        <h4>
          Streamall (netflix 'clone')
          <small class="position">Sysadmin - Backend dev - Frontend dev - Tooling</small>
        </h4>
        <p>
          In this project I designed and implemented a system that allowed hundreds of users to access thousands of long-form videos on demand, deployed on a physical cluster in a colocation with high availability and a smart caching system designed to keep 'hot' items in an SSD cache, while a massive store was on tier-3 disks.
          The project included:
          <ul>
            <li>
              Distributed transcoding of the content to a multitude of formats supported by the customer's devices.
            </li>
            <li>
              Web-ui for the administrator to handle users, API keys and content metadata
            </li>
          </ul>
        </p>
      </div>
      <div class="company-project">
        <h4>
          News parser (PYMNTS.com)
          <small class="position">Sysadmin - Backend dev - Frontend dev - Tooling</small>
        </h4>
        <p>
          The main purpose of this project was the automatization of data scraping from multiple news sources.
        </p>
        <p>
          The data is gathered by polling the sources in a parallelized manner every set amount of time, validated against a set of keywords (with associated per-source weights) and stored in a database.
        </p>
        <p>
          The sources and keywords can be managed by the users via a web interface; they also decide which matching articles are not useful and tag them as such, every month the weights and keywords get reevaluated based on ‘useless’ articles.
        </p>
        <p>
          We also have a dashboard displaying metrics, keywords, per-source matches and per-keyword matches.
        </p>
      </div>
      <div class="company-project">
        <h4>
          Multi-language interpreter to extract embedded SQL queries
        </h4>
        <p>
          We had a lot of old applications that had embedded SQL queries and the company had to migrate from Oracle 8 to Oracle 11. To achieve that, we had to analyze ALL the embedded queries so we created an parser for VB6, VB.NET, C#, ASP.NET and Java.
        </p>
      </div>
    </div>
  </section>

  <section class="experience-category">
    <h2 class="experience-title">
      Bonus tracks
    </h2>
    <div class="bonus-track">
      <p>
        Some of my personal projects. You can find more at my <a href="blog.davidventura.com.ar">blog</a>
        <ul>
          <li>
            Reverse engineered the android application to report poorly parked cars provided by the government to automate submissions of people parking in my garage
          </li>
          <li>
            Proposed a system (and proof-of-concept) to the highway authorities to catch speeding cars in freeways/highways using OpenCV to detect the license plates
          </li>
          <li>
            IoT projects:
            <ul>
              <li>Wrote a small IOT framework in MicroPython, which I use for my custom devices</li>
              <li>Used said framework to control my nightlamp with my kindle, measure ambient humidity/temperature, switch TV input sources, etc.</li>
            </ul>
          </li>
        </ul>
        <li>
          Currently writing a set of tools to handle my 3d-printer in better ways (emitting consistent gcode, cura plugin)
        </li>
        <li>
          Currently writing a video pipeline to detect and flag motion detection with my home security cameras.
        </li>
      </p>
    </div>
  </section>

  <section class="experience-category">
    <h2 class="experience-title">Lab Experience</h2>
    <div class="lab-experience">
      <p>
        Along with my normal projects I maintain a 'lab' to both host my projects and learn about new or interesting tech. Some items of note in the lab:
        <ul>
          <li>
            ARM Cluster: load-balancing LXC cluster on multiple physical ARM single-board-computers of (slightly) different architectures (armv6, AArch64) with live container migration using CRIU
          </li>
          <li>
            Jenkins for automated tests and tricky builds (one of the projects is Android-Windows-Linux)
          </li>
          <li>
            My physical router is fully deployed from the lab with ansible.
          </li>
          <li>
            All containers and VMs are fully deployed with ansible
          </li>
          <li>
            Backup system (automated ping for offline backups, daily delta and weekly snapshots uploaded to different zones, via different mechanisms)
          </li>
        </ul>
      </p>
      <p>
        All of this is monitored with generic tools (telegraf, influx, grafana, email) and also hosts some generic tooling for itself (squid, apt-cacher, git, backups)
      </p>
    </div>
  </section>

</body>
</html>