<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: I built ar-io-mlflow: Open-Source MLflow Plugin for Verifiable &amp;amp; Tamper-Proof AI Provenance in Community Articles</title>
    <link>https://community.databricks.com/t5/community-articles/i-built-ar-io-mlflow-open-source-mlflow-plugin-for-verifiable/m-p/157320#M1191</link>
    <description>&lt;P&gt;I also opened a discussion on Github, not sure which is the right place sorry if this isn't it -&amp;nbsp;&lt;A href="https://github.com/mlflow/mlflow/discussions/23355" target="_blank"&gt;https://github.com/mlflow/mlflow/discussions/23355&lt;/A&gt;&lt;/P&gt;</description>
    <pubDate>Wed, 20 May 2026 10:58:17 GMT</pubDate>
    <dc:creator>kemp</dc:creator>
    <dc:date>2026-05-20T10:58:17Z</dc:date>
    <item>
      <title>I built ar-io-mlflow: Open-Source MLflow Plugin for Verifiable &amp; Tamper-Proof AI Provenance</title>
      <link>https://community.databricks.com/t5/community-articles/i-built-ar-io-mlflow-open-source-mlflow-plugin-for-verifiable/m-p/157318#M1190</link>
      <description>&lt;P&gt;Hey everyone!&lt;/P&gt;&lt;P&gt;I've built and open-sourced ar-io-mlfow. This is a plugin that adds cryptographic provenance across the ML lifecycle (training runs, model registration, stage promotions, inference, and datasets).&lt;/P&gt;&lt;P&gt;&lt;STRONG&gt;What it does&lt;/STRONG&gt;&lt;/P&gt;&lt;UL&gt;&lt;LI&gt;Creates signed Ed25519 cryptographic proofs of your runs, models, and predictions.&lt;/LI&gt;&lt;LI&gt;Permanently anchors these small proofs (~500 bytes) to the **ar.io network** (built on Arweave decentralized storage).&lt;/LI&gt;&lt;LI&gt;Provides a `VerifiedModel` wrapper that checks integrity *before* loading/serving and raises an error on tampering.&lt;/LI&gt;&lt;LI&gt;Includes `ArioMlflowClient` (drop-in replacement) and a simple `ario_mlflow.anchor()` call.&lt;/LI&gt;&lt;LI&gt;Language-neutral verification + CLI tools for auditors.&lt;/LI&gt;&lt;/UL&gt;&lt;P&gt;The idea is to provide teams with lightweight independent verification layer to their workflows.&lt;/P&gt;&lt;P&gt;&lt;STRONG&gt;Quick Examples&lt;/STRONG&gt;&lt;/P&gt;&lt;P&gt;&lt;STRONG&gt;Training run anchoring&lt;/STRONG&gt;&lt;/P&gt;&lt;LI-CODE lang="python"&gt;import mlflow
import ario_mlflow

with mlflow.start_run():
    # ... your normal training + mlflow.log_model() ...
    result = ario_mlflow.anchor()
    print("Anchored transaction:", result["tags"]["ario.training_tx"])&lt;/LI-CODE&gt;&lt;P&gt;&lt;STRONG&gt;Verified Inference&lt;/STRONG&gt;&lt;/P&gt;&lt;LI-CODE lang="python"&gt;from ario_mlflow import VerifiedModel

vm = VerifiedModel("models:/my_model/1")   # works with Unity Catalog too
prediction = vm.predict(data)
print(prediction.proof_status)   # will raise IntegrityError if tampered&lt;/LI-CODE&gt;&lt;P&gt;Full installation, docs, architecture, and more examples are here: &lt;A href="https://github.com/ar-io/ar-io-mlflow" target="_blank" rel="noopener"&gt;https://github.com/ar-io/ar-io-mlflow&lt;/A&gt;&lt;/P&gt;&lt;P&gt;It's an early version so would love any feedback from the community so I can improve &lt;span class="lia-unicode-emoji" title=":slightly_smiling_face:"&gt;🙂&lt;/span&gt;&lt;/P&gt;&lt;P&gt;A few questions if that's ok:&lt;/P&gt;&lt;UL&gt;&lt;LI&gt;Does verifiable decentralized provenance solve a real pain point for your MLOps or governance workflows?&lt;/LI&gt;&lt;LI&gt;Any specific integration ideas or feature requests?&lt;/LI&gt;&lt;/UL&gt;&lt;P&gt;Thanks in advance if you got this far. Happy to answer any questions, review issues/PRs, or collaborate.&lt;/P&gt;&lt;P&gt;Will&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Wed, 20 May 2026 10:54:13 GMT</pubDate>
      <guid>https://community.databricks.com/t5/community-articles/i-built-ar-io-mlflow-open-source-mlflow-plugin-for-verifiable/m-p/157318#M1190</guid>
      <dc:creator>kemp</dc:creator>
      <dc:date>2026-05-20T10:54:13Z</dc:date>
    </item>
    <item>
      <title>Re: I built ar-io-mlflow: Open-Source MLflow Plugin for Verifiable &amp; Tamper-Proof AI Provenance</title>
      <link>https://community.databricks.com/t5/community-articles/i-built-ar-io-mlflow-open-source-mlflow-plugin-for-verifiable/m-p/157320#M1191</link>
      <description>&lt;P&gt;I also opened a discussion on Github, not sure which is the right place sorry if this isn't it -&amp;nbsp;&lt;A href="https://github.com/mlflow/mlflow/discussions/23355" target="_blank"&gt;https://github.com/mlflow/mlflow/discussions/23355&lt;/A&gt;&lt;/P&gt;</description>
      <pubDate>Wed, 20 May 2026 10:58:17 GMT</pubDate>
      <guid>https://community.databricks.com/t5/community-articles/i-built-ar-io-mlflow-open-source-mlflow-plugin-for-verifiable/m-p/157320#M1191</guid>
      <dc:creator>kemp</dc:creator>
      <dc:date>2026-05-20T10:58:17Z</dc:date>
    </item>
  </channel>
</rss>

