<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: nvdnadj92</title><link>https://news.ycombinator.com/user?id=nvdnadj92</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 21 Nov 2026 00:24:55 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=nvdnadj92" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by nvdnadj92 in "Pluto AI is top ranked in StarCraft BroodWar competitive scene"]]></title><description><![CDATA[
<p>Wins against top 7 Protoss player with an unconventional approach using the least used unit in the game (ghost).</p>
]]></description><pubDate>Sun, 20 Sep 2026 08:25:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=49773717</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=49773717</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49773717</guid></item><item><title><![CDATA[Pluto AI is top ranked in StarCraft BroodWar competitive scene]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.youtube.com/watch?v=1vsTqNwHquE">https://www.youtube.com/watch?v=1vsTqNwHquE</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49773716">https://news.ycombinator.com/item?id=49773716</a></p>
<p>Points: 2</p>
<p># Comments: 2</p>
]]></description><pubDate>Sun, 20 Sep 2026 08:25:54 +0000</pubDate><link>https://www.youtube.com/watch?v=1vsTqNwHquE</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=49773716</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49773716</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "When does MCP make sense vs CLI?"]]></title><description><![CDATA[
<p>One clear use case where MCP is better than anything else: design system documentation for an enterprise, or using a specific UI library. Daisy UI has an example: <a href="https://github.com/birdseyevue/daisyui-mcp" rel="nofollow">https://github.com/birdseyevue/daisyui-mcp</a><p>Context7 is a good MCP. If one were to point an agent to a docs website, the amount of tokens consumed would use up too much of the context window to be able to do a meaningfully complex task with it.<p>Figma MCP translates Figma to a language an agent understands.<p>Not everything can be a cli.</p>
]]></description><pubDate>Mon, 02 Mar 2026 06:19:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=47214494</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=47214494</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47214494</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: Who is hiring? (February 2026)"]]></title><description><![CDATA[
<p>Carta | Senior Software Engineer II | Waterloo / Toronto | ONSITE (Hybrid OK)<p>Apply Here:<p>- <a href="https://job-boards.greenhouse.io/carta/jobs/7544237003" rel="nofollow">https://job-boards.greenhouse.io/carta/jobs/7544237003</a><p>- <a href="https://job-boards.greenhouse.io/carta/jobs/7504452003" rel="nofollow">https://job-boards.greenhouse.io/carta/jobs/7504452003</a><p>⎯<p>Carta connects founders, investors, and limited partners through world-class software, purpose-built for everyone in venture capital, private equity and private credit.<p>Frontend Platform: You will own the core infrastructure and monorepo that underpins our entire frontend. Our mantra: "Create a development experience that people will miss when they leave."<p>Design Systems (Ink): You will build the foundational React components, tooling, and accessibility standards used by hundreds of Carta engineers to ship consistent UI. Check out our docs site at <a href="https://ink.carta.com" rel="nofollow">https://ink.carta.com</a><p>⎯<p>Tech: React, TypeScript, Rspack/Vite/Rush, Node.js.<p>We are looking for engineers who care deeply about developer experience, accessibility, and enabling great design at scale.</p>
]]></description><pubDate>Mon, 02 Feb 2026 20:05:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=46860683</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=46860683</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46860683</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: What Are You Working On? (Nov 2025)"]]></title><description><![CDATA[
<p>Amazing! I was planning on making a tool just like this, will take yours for a spin!</p>
]]></description><pubDate>Sun, 09 Nov 2025 23:11:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=45870202</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45870202</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45870202</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: What Are You Working On? (Nov 2025)"]]></title><description><![CDATA[
<p>Building a set of experiments that explores LLMs visual understanding of your photos to learn about you, especially given the recent learnings from deepseek-OCR. Part of the experiments delve into storing the memories with GraphRAG so they can be effectively retrieved without losing too information.</p>
]]></description><pubDate>Sun, 09 Nov 2025 22:52:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=45870041</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45870041</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45870041</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: Who uses open LLMs and coding assistants locally? Share setup and laptop"]]></title><description><![CDATA[
<p>Laptop: Apple M2 Max, 32GB memory (2023)<p>Setup:<p>Terminal:<p>- Ghostty + Starship for modern terminal experience<p>- Homebrew to install system packages<p>IDE:<p>- Zed (can connect to local models via LM-Studio server)<p>- also experimenting with warp.dev<p>LLMs:<p>- LM-studio as open-source model playground<p>- GPT-OSS 20B<p>- QWEN3-Coder-30B-AEB-quantized-4bit<p>- Gemma3-12B<p>Other utilities:<p>- Rectangle.app (window tile manager)<p>- Wispr.flow - create voice notes<p>- Obsidian - track markdown notes</p>
]]></description><pubDate>Sat, 01 Nov 2025 00:01:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=45778040</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45778040</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45778040</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Story Keeper – AI agents with narrative continuity instead of memory"]]></title><description><![CDATA[
<p>After reading this post and the readme, I'm not convinced that this is solving a real, observed problem. You outline an example with the long-term coaching mentorship, but why or how is your solution preferable to telling Claude to maintain a set of notes and observations about you, similar to <a href="https://github.com/heyitsnoah/claudesidian" rel="nofollow">https://github.com/heyitsnoah/claudesidian</a>?<p>the jazz metaphors do not help provide additional context.</p>
]]></description><pubDate>Thu, 30 Oct 2025 22:39:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=45766232</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45766232</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45766232</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Chonky – a neural text semantic chunking goes multilingual"]]></title><description><![CDATA[
<p>In case you need conversational data for the experiment you want to try, I developed an open-source cli tool [1] that create transcripts from voice chats on discord. Feel free to try it out!<p>[1] <a href="https://github.com/naveedn/audio-transcriber" rel="nofollow">https://github.com/naveedn/audio-transcriber</a></p>
]]></description><pubDate>Sun, 26 Oct 2025 16:33:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=45713136</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45713136</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45713136</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Your data model is your destiny"]]></title><description><![CDATA[
<p>Agree with the first half of the article, but every example the author pointed out predates AI. What are examples of companies that have been founded in the past 3 years and prove the authors point that the data model is the definitive edge?</p>
]]></description><pubDate>Thu, 16 Oct 2025 22:55:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=45611680</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45611680</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45611680</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Scriber Pro – Offline AI transcription for macOS"]]></title><description><![CDATA[
<p>Yes, but by a negligible margin. My program is designed for multi-track audio, which means I run this in parallel on multiple 3 hour recordings, and get results in 12 minutes.<p>You haven’t shared any architectural details. What model? What size? How can anyone be sure that what you’re building is truly offline?</p>
]]></description><pubDate>Wed, 15 Oct 2025 13:51:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=45592574</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45592574</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45592574</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Scriber Pro – Offline AI transcription for macOS"]]></title><description><![CDATA[
<p>I vibecoded a similar app. Here’s the open source link, if folks want to build their own:<p><a href="https://github.com/naveedn/audio-transcriber" rel="nofollow">https://github.com/naveedn/audio-transcriber</a></p>
]]></description><pubDate>Wed, 15 Oct 2025 13:00:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=45591771</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45591771</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45591771</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: What are you working on? (October 2025)"]]></title><description><![CDATA[
<p>Ah lovely! I’d be happy to assist, create an issue on GitHub and we can go from there!</p>
]]></description><pubDate>Mon, 13 Oct 2025 02:53:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=45564271</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45564271</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45564271</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Ask HN: What are you working on? (October 2025)"]]></title><description><![CDATA[
<p>I wanted to build my own speech-to-text transcription program [1] for Discord, similar to how zoom or google hangouts works. I built it so that I can record my group's DND sessions and build applications / tools for VTTs (Virtual TableTop gaming).<p>It can process a set of 3-hour audio files in ~20 mins.<p>I recorded a demo video of how it works here: <a href="https://www.youtube.com/watch?v=v0KZGyJARts&t=300s" rel="nofollow">https://www.youtube.com/watch?v=v0KZGyJARts&t=300s</a><p>[1] <a href="https://github.com/naveedn/audio-transcriber" rel="nofollow">https://github.com/naveedn/audio-transcriber</a><p>I alluded to building this tool on a previous HN thread: <a href="https://news.ycombinator.com/item?id=45338694">https://news.ycombinator.com/item?id=45338694</a></p>
]]></description><pubDate>Mon, 13 Oct 2025 00:10:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=45563328</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45563328</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45563328</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Python Audio Transcription: Convert Speech to Text Locally"]]></title><description><![CDATA[
<p>I would suggest 2 speaker-diarization libraries:<p>- <a href="https://huggingface.co/pyannote/speaker-diarization-3.1" rel="nofollow">https://huggingface.co/pyannote/speaker-diarization-3.1</a>
- <a href="https://github.com/narcotic-sh/senko" rel="nofollow">https://github.com/narcotic-sh/senko</a><p>I personally love senko since it can run in seconds, whereas py-annote took hours, but there is a 10% WER (word error rate) that is tough to get around.</p>
]]></description><pubDate>Mon, 22 Sep 2025 20:28:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=45339017</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45339017</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45339017</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Show HN: Python Audio Transcription: Convert Speech to Text Locally"]]></title><description><![CDATA[
<p>I'm working on the same project myself and was planning to write a blog post similar to the author's. However, I'll share some additional tips and tricks that really made a difference for me.<p>For preprocessing, I found it best to convert files to a 16kHz WAV format for optimal processing. I also add low-pass and high-pass filters to remove non-speech sounds. To avoid hallucinations, I run Silero VAD on the entire audio file to find timestamps where there's a speaker. A side note on this: Silero requires careful tuning to prevent audio segments from being chopped up and clipped. I also use a post-processing step to merge adjacent VAD chunks, which helps ensure cohesive Whisper recordings.<p>For the Whisper task, I run Whisper in small audio chunks that correspond to the VAD timestamps. Otherwise, it will hallucinate during silences and regurgitate the passed-in prompt. If you're on a Mac, use the whisper-mlx models from Hugging Face to speed up transcription. I ran a performance benchmark, and it made a 22x difference to use a model designed for the Apple Neural Engine.<p>For post-processing, I've found that running the generated SRT files through ChatGPT to identify and remove hallucination chunks has a better yield.</p>
]]></description><pubDate>Mon, 22 Sep 2025 19:59:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=45338694</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=45338694</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45338694</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "A distributed systems reading list"]]></title><description><![CDATA[
<p>The best introductory resource I have found in my career was: "Distributed Systems for Fun and Profit" by Mixu. It's about 50 pages long, and is broken down quite well.<p><a href="https://book.mixu.net/distsys/single-page.html" rel="nofollow">https://book.mixu.net/distsys/single-page.html</a></p>
]]></description><pubDate>Thu, 08 Feb 2024 20:33:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=39307301</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=39307301</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39307301</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Carta lays off 10% of the workforce"]]></title><description><![CDATA[
<p>Been at the company 3 years, this guy Manu was there in the early days, but he is pretty much a ghost since the past 7 years. He's moved on, I think he is a VC now.</p>
]]></description><pubDate>Wed, 11 Jan 2023 17:44:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=34341562</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=34341562</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34341562</guid></item><item><title><![CDATA[New comment by nvdnadj92 in "Carta lays off 10% of the workforce"]]></title><description><![CDATA[
<p>better than nothing, in my opinion. At least they were considered.</p>
]]></description><pubDate>Wed, 11 Jan 2023 17:43:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=34341543</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=34341543</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34341543</guid></item><item><title><![CDATA[Carta lays off 10% of the workforce]]></title><description><![CDATA[
<p>Just got off company announcement, here are the key details:<p><pre><code>   - the departure affected 10% of the company
</code></pre>
The benefits include:<p><pre><code>   - 2.5 months of severance, plus one week for every year of tenure
   - 3 months of health insurance
   - 1 year cliff for RSUs waived for new employees. In cases where employees did not get equity, they get a lump sum payout
   - Visa holders get one 30-minute paid consultation with an immigration attorney
   - Voluntary self departures accepted until 12PM tomorrow, but self-departures only get 2.5 months (tenure not factored in severance)</code></pre></p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=34341186">https://news.ycombinator.com/item?id=34341186</a></p>
<p>Points: 70</p>
<p># Comments: 85</p>
]]></description><pubDate>Wed, 11 Jan 2023 17:24:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=34341186</link><dc:creator>nvdnadj92</dc:creator><comments>https://news.ycombinator.com/item?id=34341186</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34341186</guid></item></channel></rss>