Showcase

ML/AI engineer portfolio website, built by Underdog Saluki 27B

IQ2_XSneeds 10.2 GB at 32Kvia OpenCodellama.cpp
Running live in your browserNew tab
Memory to run
10.2GB
At 32K context, from its benchmark run
Show the code the model wrote
HTML8,567 characters
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Mira Kovač — Senior ML Engineer</title>
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link href="https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@400;500;600&display=swap" rel="stylesheet">
<style>
  :root {
    --bg: #0a0a0a;
    --fg: #d4d4d4;
    --bright: #f2f2f2;
    --muted: #8a8a8a;
    --line: #2a2a2a;
    --green: #33ff55;
    --space-1: 8px;
    --space-2: 16px;
    --space-3: 24px;
    --space-4: 40px;
    --space-5: 64px;
    --max: 720px;
  }

  html { color-scheme: dark; }
  body {
    margin: 0;
    background: var(--bg);
    color: var(--fg);
    font-family: "JetBrains Mono", ui-monospace, SFMono-Regular, Menlo, Consolas, monospace;
    font-size: 15px;
    line-height: 1.55;
    -webkit-font-smoothing: antialiased;
  }
  ::selection { background: rgba(51, 255, 85, 0.25); color: var(--bright); }

  .container {
    max-width: var(--max);
    margin: 0 auto;
    padding: var(--space-5) var(--space-3);
  }

  a {
    color: var(--green);
    text-decoration: none;
  }
  a:hover { text-decoration: underline; }
  a:focus-visible { outline: 1px solid var(--green); outline-offset: 3px; }

  /* header */
  .prompt {
    color: var(--muted);
    font-size: 13px;
    margin: 0 0 var(--space-2);
  }
  h1 {
    color: var(--bright);
    font-weight: 600;
    font-size: 34px;
    letter-spacing: -0.5px;
    margin: 0 0 var(--space-1);
  }
  .subtitle {
    color: var(--muted);
    font-size: 15px;
    margin: 0;
  }
  .cursor {
    color: var(--green);
    animation: blink 1s step-end infinite;
  }
  @keyframes blink { 50% { opacity: 0; } }
  @media (prefers-reduced-motion: reduce) {
    .cursor { animation: none; }
  }

  /* sections */
  section { margin-top: var(--space-5); }
  h2 {
    font-size: 13px;
    font-weight: 600;
    text-transform: uppercase;
    letter-spacing: 0.12em;
    color: var(--bright);
    margin: 0 0 var(--space-3);
  }
  h2::before {
    content: "// ";
    color: var(--muted);
  }

  /* experience timeline */
  .timeline {
    list-style: none;
    margin: 0;
    padding: 0 0 0 var(--space-3);
    border-left: 1px solid var(--line);
    margin-left: var(--space-1);
  }
  .timeline > li {
    position: relative;
    margin-bottom: var(--space-4);
  }
  .timeline > li:last-child { margin-bottom: 0; }
  .timeline > li::before {
    content: "";
    position: absolute;
    left: calc(-1 * var(--space-3) - 5px);
    top: 7px;
    width: 9px;
    height: 9px;
    background: var(--green);
  }
  .year {
    color: var(--green);
    font-size: 13px;
    letter-spacing: 0.08em;
    margin: 0 0 var(--space-1);
  }
  h3 {
    color: var(--bright);
    font-weight: 600;
    font-size: 17px;
    margin: 0;
  }
  .role {
    color: var(--muted);
    font-size: 14px;
    margin: 0 0 var(--space-2);
  }
  .bullets {
    list-style: none;
    margin: 0;
    padding: 0;
  }
  .bullets li {
    position: relative;
    padding-left: var(--space-2);
    margin-top: var(--space-1);
  }
  .bullets li::before {
    content: ">";
    position: absolute;
    left: 0;
    color: var(--muted);
  }

  /* skills */
  .chips {
    list-style: none;
    margin: 0;
    padding: 0;
    display: flex;
    flex-wrap: wrap;
    gap: var(--space-1);
  }
  .chips li {
    border: 1px solid var(--green);
    padding: 6px 12px;
    font-size: 14px;
    color: var(--fg);
  }

  /* projects */
  .projects {
    list-style: none;
    margin: 0;
    padding: 0;
  }
  .projects li {
    margin-top: var(--space-2);
  }
  .projects li:first-child { margin-top: 0; }
  .projects .name {
    color: var(--bright);
    font-weight: 600;
  }
  .projects .desc {
    color: var(--muted);
  }

  /* publication */
  .pub { margin: 0; }
  .pub .authors { color: var(--bright); font-weight: 600; }
  .pub .title { color: var(--fg); }
  .pub .venue { color: var(--muted); font-size: 14px; }

  /* contact */
  .contact { margin: 0; font-size: 14px; }
  .contact .sep { color: var(--muted); }

  footer {
    margin-top: var(--space-5);
    color: var(--muted);
    font-size: 13px;
  }

  @media (max-width: 600px) {
    .container { padding: var(--space-4) var(--space-2); }
    body { font-size: 14px; }
    h1 { font-size: 28px; }
    section { margin-top: var(--space-4); }
    footer { margin-top: var(--space-4); }
  }
</style>
</head>
<body>
<div class="container">

  <header>
    <p class="prompt">mira@berlin:~$ whoami</p>
    <h1>Mira Kovač<span class="cursor" aria-hidden="true">█</span></h1>
    <p class="subtitle">Senior Machine Learning Engineer · Berlin</p>
  </header>

  <section aria-labelledby="exp-h">
    <h2 id="exp-h">Experience</h2>
    <ol class="timeline">
      <li>
        <p class="year">2023 — PRESENT</p>
        <h3>Helsing</h3>
        <p class="role">Senior Machine Learning Engineer · Berlin</p>
        <ul class="bullets">
          <li>Rebuilt the credit-risk scoring model from a single logistic regression to a gradient-boosted ensemble, cutting annualized loss by 1.4 pts while holding calibration within 1%.</li>
          <li>Shipped a near-real-time feature store (Kafka + Redis) feeding 40+ features into the risk pipeline, reducing decision latency from 400 ms to 60 ms.</li>
          <li>Owns the end-to-end MLOps pipeline — training, evaluation, canary deployment — for the six models in production.</li>
        </ul>
      </li>
      <li>
        <p class="year">2020 — 2023</p>
        <h3>Zalando</h3>
        <p class="role">Machine Learning Engineer · Berlin</p>
        <ul class="bullets">
          <li>Built a two-tower retrieval + fine-ranking stack for on-site search, improving click-through rate by 6% over the previous linear baseline.</li>
          <li>Deployed a near-real-time CTR prediction pipeline handling 2M requests/day with p95 latency under 50 ms.</li>
          <li>Cut inference cost by 30% via INT8 quantization and ONNX export of the ranking model.</li>
        </ul>
      </li>
      <li>
        <p class="year">2017 — 2020</p>
        <h3>Deutsche Telekom</h3>
        <p class="role">Data Scientist · Munich</p>
        <ul class="bullets">
          <li>Developed churn-prediction models for mobile subscribers, reducing churn by 3% among the top decile of the risk score.</li>
          <li>Built a classification pipeline routing 1.2M support tickets/month into 40 intent categories, with 92% of tickets reaching the right team.</li>
          <li>Wrote the team's A/B testing playbook and ran 30+ experiments on pricing and retention.</li>
        </ul>
      </li>
    </ol>
  </section>

  <section aria-labelledby="skills-h">
    <h2 id="skills-h">Skills</h2>
    <ul class="chips">
      <li>Python</li>
      <li>PyTorch</li>
      <li>TensorFlow</li>
      <li>SQL</li>
      <li>Kafka</li>
      <li>Kubernetes</li>
      <li>MLflow</li>
    </ul>
  </section>

  <section aria-labelledby="projects-h">
    <h2 id="projects-h">Side Projects</h2>
    <ul class="projects">
      <li><span class="name">gridwalk</span> <span class="desc">— a Rust RL agent for 2D-grid pathfinding, built to compare PPO against hand-rolled value iteration.</span></li>
      <li><span class="name">papertrail</span> <span class="desc">— a self-hosted arXiv + BibTeX tracker with a terminal UI: one key to search, one to cite.</span></li>
      <li><span class="name">noise-floor</span> <span class="desc">— a time-series anomaly detector for home IoT sensors, trained on 18 months of readings.</span></li>
    </ul>
  </section>

  <section aria-labelledby="pub-h">
    <h2 id="pub-h">Publication</h2>
    <p class="pub">
      <span class="authors">Kovač, M. &amp; Richter, S.</span>
      <span class="title">"When Trees Beat Networks: A Systematic Comparison of Gradient-Boosted and Neural Models for Tabular Credit-Risk Data."</span>
      <span class="venue">Journal of Machine Learning Research, 25(12), 2024.</span>
    </p>
  </section>

  <section aria-labelledby="contact-h">
    <h2 id="contact-h">Contact</h2>
    <p class="contact">
      <a href="mailto:mira.kovac@example.com">mira.kovac@example.com</a>
      <span class="sep"> · </span>
      <a href="https://github.com/mirakovac">github.com/mirakovac</a>
      <span class="sep"> · </span>
      <a href="https://linkedin.com/in/mirakovac">linkedin.com/in/mirakovac</a>
    </p>
  </section>

  <footer>exit 0</footer>

</div>
</body>
</html>

Same prompt, other models

8 more
Compare side by side

One email when a new model lands on the board or in the showcase. Unsubscribe anytime.