index.html

<!doctype html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Programme World Model · Bruno Chikeka</title>
  <meta name="description" content="Evidence-grounded AI for complex programmes. A research and engineering project by Bruno Chikeka.">
  <style>
    :root{color-scheme:light;--ink:#142e33;--muted:#506568;--paper:#f5f3ec;--line:#d4dcd5;--accent:#17675d}
    *{box-sizing:border-box}body{margin:0;background:var(--paper);color:var(--ink);font:17px/1.65 system-ui,-apple-system,BlinkMacSystemFont,"Segoe UI",sans-serif}
    main{max-width:1120px;margin:auto;padding:36px 40px 60px}a{color:var(--accent);text-underline-offset:4px}a:hover{text-decoration-thickness:2px}a:focus-visible{outline:3px solid var(--accent);outline-offset:5px}
    .masthead{display:flex;justify-content:space-between;gap:20px;padding-bottom:28px;border-bottom:1px solid var(--line);font-size:14px}.masthead strong{letter-spacing:.02em}.masthead a{color:var(--ink)}
    .hero{padding:60px 0 42px;max-width:930px}.eyebrow{font-size:12px;text-transform:uppercase;letter-spacing:.15em;font-weight:700;color:var(--accent)}h1{font:normal clamp(42px,7vw,78px)/1.06 Georgia,"Times New Roman",serif;letter-spacing:-.035em;margin:18px 0 25px}h1 span{display:block}h2{font:normal 32px/1.2 Georgia,"Times New Roman",serif;margin:0 0 18px}h3{font-size:17px;margin:0 0 10px}.lead{font-size:21px;max-width:740px;color:var(--muted)}.status{display:inline-block;border:1px solid #b1c7bb;border-radius:4px;padding:7px 12px;font-size:13px;margin:14px 0 0;background:#e9eee5}
    .metrics{display:grid;grid-template-columns:repeat(3,1fr);border-top:1px solid var(--line);border-bottom:1px solid var(--line);padding:24px 0;gap:20px}.metric b{display:block;font-size:34px;font-weight:500;line-height:1.2}.metric span{font-size:14px;color:var(--muted)}
    section{padding:42px 0;border-bottom:1px solid var(--line)}.split{display:grid;grid-template-columns:1fr 2fr;gap:52px}.split p:first-child{margin-top:0}.split p:last-child{margin-bottom:0}.columns{display:grid;grid-template-columns:repeat(3,1fr);gap:28px}.columns p{font-size:15px;color:var(--muted);margin-bottom:0}.step{display:block;font:13px ui-monospace,monospace;color:var(--accent);margin-bottom:13px}.note{color:var(--muted);font-size:14px}.finding{padding:22px 26px;border-left:3px solid var(--accent);background:#e9eee5}.finding p{margin:0}.flow{display:flex;gap:12px;align-items:center;flex-wrap:wrap;margin-top:28px;font-size:14px}.flow span{padding:10px 14px;border:1px solid var(--line);border-radius:3px}.flow .active{background:var(--ink);color:white;border-color:var(--ink)}.flow i{font-style:normal;color:var(--muted)}footer{display:flex;justify-content:space-between;gap:20px;padding-top:26px;font-size:13px;color:var(--muted)}
    @media(max-width:650px){main{padding:24px 22px 36px}.hero{padding:38px 0 30px}.lead{font-size:18px}.split,.columns{grid-template-columns:1fr;gap:20px}.metrics{gap:12px}.metric b{font-size:29px}.metric span{font-size:12px}section{padding:30px 0}footer{flex-direction:column;gap:8px}.masthead{font-size:12px}}
  </style>
</head>
<body>
<main>
  <header class="masthead"><strong>Bruno Chikeka</strong><a href="https://huggingface.co/Chikeka" target="_blank" rel="noopener">Research &amp; engineering ↗</a></header>
  <div class="hero">
    <div class="eyebrow">Evidence-grounded AI · Major programmes</div>
    <h1>Programme<span>World Model.</span></h1>
    <p class="lead">Understanding how complex programmes change, with a traceable path from the source report to the model’s view of the world.</p>
    <p class="status">Research prototype · Evidence pipeline and synthetic kernel</p>
  </div>
  <div class="metrics" aria-label="Current pilot coverage">
    <div class="metric"><b>15</b><span>original project reports preserved</span></div>
    <div class="metric"><b>135</b><span>source-linked project cells</span></div>
    <div class="metric"><b>60</b><span>automated software tests passing</span></div>
  </div>
  <p class="note">Pilot snapshot · 9 September 2026 · One World Bank project, P126507. These are engineering and coverage measures, not predictive accuracy scores.</p>
  <section class="split">
    <h2>From reports<br>to programme histories.</h2>
    <div><p>Programme delivery leaves a fragmented record: implementation reports, revised dates, ratings and decisions. Before a model can learn from that record, we need to establish what was reported, which entity it concerns, and when it could have been known.</p><p>This project begins there. It preserves original documents, extracts cited observations and reconstructs differences between available reports. The longer-term aim is to explore dependencies, test conditional scenarios and update programme state as evidence changes.</p></div>
  </section>
  <section>
    <h2>Three foundations.</h2>
    <div class="columns">
      <div><span class="step">01 / EVIDENCE</span><h3>Every observation has a source.</h3><p>Original bytes and hashes, report identities, exact excerpts and PDF-page references. Missing information remains unknown.</p></div>
      <div><span class="step">02 / TIME</span><h3>Publication is not the event.</h3><p>Report dates, disclosure metadata and retrieval times stay separate. Historical availability requires its own verification.</p></div>
      <div><span class="step">03 / VALIDATION</span><h3>Agreement is not ground truth.</h3><p>Versioned evidence, independent-review packets and explicit reconciliation. Observations and synthetic assumptions remain distinguishable.</p></div>
    </div>
  </section>
  <section class="split">
    <h2>A finding that<br>changes the question.</h2>
    <div class="finding"><p>In the pilot, a project summary’s revised closing date differs from the closing date of its IDA credit. The credit and a trust fund have separate dates. A changed summary cell therefore cannot automatically become a label for a legal extension of the same loan.</p><p class="note" style="margin-top:14px">This is why entity scope and source review come before model training.</p></div>
  </section>
  <section class="split">
    <h2>What exists today.</h2>
    <div><p>A reproducible public-report evidence pipeline, a versioned SQLite index, resumable acquisition, review and reconciliation records, backup recovery, and automated checks on Python 3.11 and 3.14.</p><p>A separate synthetic simulation kernel supports testing explicitly assumed mechanisms. It is not empirically calibrated.</p><p class="note">The pilot contains 121 reported values and 14 explicit unknowns. Sequence 7 is missing from the acquired series. Independent extraction accuracy, historical source versions and financing-specific outcome definitions remain open.</p></div>
  </section>
  <section>
    <h2>The research path.</h2>
    <div class="flow" aria-label="Research stages"><span class="active">Evidence &amp; provenance</span><i aria-hidden="true">→</i><span>Independent validation</span><i aria-hidden="true">→</i><span>Empirical forecasting</span></div>
    <p class="note">There is no trained programme predictor yet. Future datasets and experiments will be shared with their provenance, evaluation and limitations when ready. This Space is a project overview, not an interactive forecasting tool.</p>
    <p><a href="https://huggingface.co/Chikeka" target="_blank" rel="noopener">Follow the work on Hugging Face ↗</a></p>
  </section>
  <footer><span>Programme World Model · Bruno Chikeka</span><span>Traceable evidence. Explicit assumptions. Reproducible evaluation.</span></footer>
</main>
</body>
</html>