<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Aman Sharma</title><link>https://amansharma612.github.io/</link><description>Recent content on Aman Sharma</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 25 Jul 2026 19:25:00 +0000</lastBuildDate><atom:link href="https://amansharma612.github.io/index.xml" rel="self" type="application/rss+xml"/><item><title>Decoding DSpark</title><link>https://amansharma612.github.io/posts/2026-07-26-dspark/</link><pubDate>Sat, 25 Jul 2026 19:25:00 +0000</pubDate><guid>https://amansharma612.github.io/posts/2026-07-26-dspark/</guid><description>The ins and outs of speculative decoding</description></item><item><title>About</title><link>https://amansharma612.github.io/about/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://amansharma612.github.io/about/</guid><description>&lt;p&gt;Hello there! I am currently an SWE @ Agrani Labs. I graduated from IIT Bombay in 2025.&lt;/p&gt;
&lt;p&gt;On the technical side, I am deeply passionate towards the areas of architecture, distributed systems, compilers and LLM training &amp;amp; inference. In general, anything technical deeply intrigues me and usually I am not satisfied till I understand&lt;br&gt;
it to its fundamental concepts.&lt;/p&gt;
&lt;p&gt;I am currently working on/exploring compilation for heterogenous hardware.&lt;/p&gt;
&lt;p&gt;Previously, I have had experience in Natural Language Processing and Information Retrieval through my coursework. You can visit the &amp;ldquo;Projects&amp;rdquo; page to know more.&lt;/p&gt;</description></item><item><title>Data Adaptive Hashing for Document Retreival</title><link>https://amansharma612.github.io/projects/4_project/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://amansharma612.github.io/projects/4_project/</guid><description>Exploring LSH and Neural LSH algorithms for dense embedding search</description></item><item><title>Document and Query Likelihood Evaluations for Learning to Rank Systems</title><link>https://amansharma612.github.io/projects/3_project/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://amansharma612.github.io/projects/3_project/</guid><description>Assessing Quality of Neural Re-ranking based on certain training objectives</description></item><item><title>Micrograd CUDA - Autodiff Tensor Compute Engine</title><link>https://amansharma612.github.io/projects/5_project/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://amansharma612.github.io/projects/5_project/</guid><description>Tensor compute engine for NVIDIA GPUs</description></item></channel></rss>