<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Blog — Eryawan Presma Yulianrifat</title>
    <link>https://eryawan.com/blog/</link>
    <description>17 long-form posts: machine learning surveys, retrieval, reinforcement learning, speech, software practice, and startup economics in Indonesia.</description>
    <language>en</language>
    <lastBuildDate>Sat, 31 Oct 2026 00:00:00 GMT</lastBuildDate>
    <atom:link href="https://eryawan.com/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Software Will Eat Services by Buying Them</title>
      <link>https://eryawan.com/blog/software-will-eat-services/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/software-will-eat-services/</guid>
      <pubDate>Sat, 31 Oct 2026 00:00:00 GMT</pubDate>
      <description>Why the next software companies will buy the service firms they automate: the value equation, where the moat lives, and what breaks it.</description>
      <category>economics</category>
      <category>private-equity</category>
      <category>indonesia</category>
      <category>startup</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Retrieval Encoder Training Operationalization</title>
      <link>https://eryawan.com/blog/encoder-training-operationalization/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/encoder-training-operationalization/</guid>
      <pubDate>Sat, 30 May 2026 00:00:00 GMT</pubDate>
      <description>Decide if a retrieval encoder is worth training over BM25 or an API model on nDCG@10, latency and index size, then train one.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>information-retrieval</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Recommender Systems - Item Neighbors to Generative Models</title>
      <link>https://eryawan.com/blog/recommender-systems-arc/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/recommender-systems-arc/</guid>
      <pubDate>Sat, 30 May 2026 00:00:00 GMT</pubDate>
      <description>Recommender systems from Amazon&apos;s item neighbors and the Netflix Prize to two-tower retrieval and generative semantic IDs.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>recommender-systems</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>A Vibecoder&apos;s Survival Guide to Software Testing</title>
      <link>https://eryawan.com/blog/software-testing-survival-guide/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/software-testing-survival-guide/</guid>
      <pubDate>Sat, 30 May 2026 00:00:00 GMT</pubDate>
      <description>Software testing one kind per section, unit to chaos to LLM evals: a tiny example, the bug it catches, and the tool to use.</description>
      <category>software-engineering</category>
      <category>testing</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Encoders - Squash Reality into a Vector Space</title>
      <link>https://eryawan.com/blog/encoders-squash-reality/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/encoders-squash-reality/</guid>
      <pubDate>Wed, 27 May 2026 00:00:00 GMT</pubDate>
      <description>How encoders squash images and text into vectors, from SIFT and word2vec through BERT and ViT to CLIP&apos;s shared space.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Why Innovation Does Not Live in Indonesia</title>
      <link>https://eryawan.com/blog/innovation-indonesia/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/innovation-indonesia/</guid>
      <pubDate>Wed, 27 May 2026 00:00:00 GMT</pubDate>
      <description>Why Indonesia&apos;s innovation flywheel will not spin: rent-seeking grants, a trust deficit, no exit path, and where to push.</description>
      <category>economics</category>
      <category>innovation</category>
      <category>indonesia</category>
      <category>startup</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
      <dc:creator>Bryan Jeshua</dc:creator>
    </item>
    <item>
      <title>Agentic AI - From Symbolic Planning to Autonomous Agents</title>
      <link>https://eryawan.com/blog/agentic-harness/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/agentic-harness/</guid>
      <pubDate>Tue, 19 May 2026 00:00:00 GMT</pubDate>
      <description>Agentic AI from STRIPS and BDI to ReAct, Toolformer, GRPO and MCP, with open problems and seven portfolio projects.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Model2Vec, A Static-Embedding Distillation Algorithm, Explained</title>
      <link>https://eryawan.com/blog/model2vec-explained/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/model2vec-explained/</guid>
      <pubDate>Tue, 19 May 2026 00:00:00 GMT</pubDate>
      <description>Compress a sentence transformer into a CPU lookup table: vocab prep, PCA, SIF weighting, quantization, and a ~50-line reference.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Representation Analysis, A Reading Path</title>
      <link>https://eryawan.com/blog/representation-analysis-reading-path/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/representation-analysis-reading-path/</guid>
      <pubDate>Tue, 19 May 2026 00:00:00 GMT</pubDate>
      <description>Representation analysis in eight papers, from Mikolov&apos;s 2013 analogies to sparse autoencoders: concepts as directions.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>World Models - From Dyna to Foundation Simulators</title>
      <link>https://eryawan.com/blog/world-models/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/world-models/</guid>
      <pubDate>Tue, 19 May 2026 00:00:00 GMT</pubDate>
      <description>World models from Sutton&apos;s 1990 Dyna through Dreamer, Genie and video simulators to LeCun&apos;s JEPA, with open problems.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Spoken Language Processing</title>
      <link>https://eryawan.com/blog/speech-processing/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/speech-processing/</guid>
      <pubDate>Sat, 11 Oct 2025 00:00:00 GMT</pubDate>
      <description>Spoken language processing notes: ASR, text-to-speech, dialog systems, phonetics, and signal sampling, Nyquist and aliasing.</description>
      <category>philosophy</category>
      <category>machine-learning</category>
      <category>math</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
      <dc:creator>Dr. Kurniawati Azizah, S.T., M.Phil. (Lecturer)</dc:creator>
    </item>
    <item>
      <title>Personal Identity Problem and Mutability Immutability</title>
      <link>https://eryawan.com/blog/personal-identity-problem/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/personal-identity-problem/</guid>
      <pubDate>Sun, 17 Aug 2025 00:00:00 GMT</pubDate>
      <description>The personal identity problem in code: the Ship of Theseus, mutable vs immutable objects, Python&apos;s id(), and primary keys.</description>
      <category>philosophy</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Why Python is Irreplaceable for Data Science, AI and Automation</title>
      <link>https://eryawan.com/blog/why-python-is-currently-irreplacable/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/why-python-is-currently-irreplacable/</guid>
      <pubDate>Tue, 15 Jul 2025 00:00:00 GMT</pubDate>
      <description>Why Python&apos;s design is hard to replace for data science and AI: C/C++ bindings, duck typing, monkey patching, introspection.</description>
      <category>python</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Collection of Foundational Math Problems</title>
      <link>https://eryawan.com/blog/math-problems/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/math-problems/</guid>
      <pubDate>Thu, 10 Apr 2025 00:00:00 GMT</pubDate>
      <description>Foundational math problems for ML: running mean and variance, linear algebra, maximum likelihood and decision theory.</description>
      <category>math</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Deep Learning Notes from Zero to Hero</title>
      <link>https://eryawan.com/blog/deep-learning/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/deep-learning/</guid>
      <pubDate>Mon, 24 Mar 2025 00:00:00 GMT</pubDate>
      <description>Deep learning course notes: probability, likelihood, information theory, regression, classification and backpropagation.</description>
      <category>machine-learning</category>
      <category>deep-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
    <item>
      <title>Reinforcement Learning Notes: Theory and Algorithms</title>
      <link>https://eryawan.com/blog/rl/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/rl/</guid>
      <pubDate>Mon, 24 Mar 2025 00:00:00 GMT</pubDate>
      <description>Reinforcement learning course notes: MDPs, Bellman equations, policy iteration, value approximation and policy gradients.</description>
      <category>machine-learning</category>
      <category>reinforcement-learning</category>
      <category>notes</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
      <dc:creator>Vektor Dewanto, S.T., M.Eng., Ph.D. (Lecturer)</dc:creator>
    </item>
    <item>
      <title>Paradigm of Making Machine Learn</title>
      <link>https://eryawan.com/blog/general-ml/</link>
      <guid isPermaLink="true">https://eryawan.com/blog/general-ml/</guid>
      <pubDate>Sat, 22 Mar 2025 00:00:00 GMT</pubDate>
      <description>Machine learning from first principles: Mitchell&apos;s definition, three paradigms by analogy, and supervised learning formalized.</description>
      <category>machine-learning</category>
      <dc:creator>Eryawan Presma Yulianrifat</dc:creator>
    </item>
  </channel>
</rss>
