i originally got into computers through gaming, starting with autohotkey macros and basic windows optimization scripts. eventually, i realized that a lot of the so-called performance tweaks people shared weren't actually doing much, which pushed me to understand what was happening underneath.

that led me into more advanced tweaking communities, experimenting with tools like scewin, hyyote-in-depth-tweaking, and calyptos-latency-guide, alongside motherboard firmware configurations, memory tuning, overclocking, and latency optimization. over time, i became more interested in understanding how hardware and operating systems worked.

eventually, my interests shifted toward linux. i started with linux mint before quickly moving to arch, then experimented with gentoo, void, artix, cachyos, nixos, bedrock, bsd, and other operating systems.

while exploring linux, i also joined an anti-extortion group focused on protecting people targeted by harmful online communities, including groups associated with roblox and 764. through this work, i gained experience with open-source intelligence (osint), operational security (opsec), threat modeling, database/databreach research, web crawling, and investigating information across websites and publicly available sources.

much of that work involved researching, verifying, and documenting information before passing relevant findings to authorities so they could handle the situation appropriately. it gave me a deeper interest in cybersecurity, digital investigations, and understanding how information is exposed and protected online.

my interest in local ai started with the belief that artificial intelligence shouldn't be locked behind ridiculous subscription prices, geographical restrictions, censorship, or corporate control. access to powerful technology and information shouldn't depend on someone's financial situation, where they live, or what they're permitted to access. i believe independent journalists, researchers, and everyday people should have access to tools that help them investigate, learn, question narratives, and explore information freely. technology should empower people to think independently rather than decide what they're allowed to know. privacy is just as important to me. people should have control over their own data, conversations, research, and software without being forced to rely on centralized services or surrender sensitive information to third parties. these beliefs are what pushed me toward open-weight models, local inference, llama.cpp, cuda optimization, and running language models on my own hardware. i'm interested in making ai more accessible, affordable, private, and independent.

one of my long-term goals is to start a company focused on making powerful local ai accessible to everyone, regardless of their hardware or financial situation. i want to research ways to achieve the reasoning and intelligence of much larger models, such as 30b-parameter models, while operating within the computational and memory requirements of something closer to a 7b model. rather than relying solely on quantization, which can introduce quality loss, i'm especially interested in optimizing token processing, context management, inference i/o, and how information is delivered to and processed by language models. my goal is to preserve as much of the larger model's capabilities as possible while dramatically reducing hardware requirements, memory consumption, and inference costs. i believe privacy shouldn't require expensive hardware, and powerful ai shouldn't be reserved for people who can afford dedicated machines costing thousands of dollars. i want to help make private, locally hosted intelligence accessible to ordinary people, independent journalists, researchers, and developers without forcing them to depend on expensive cloud services.

nowadays, i spend most of my time learning and experimenting with programming, reverse engineering, low-level software development, and local artificial intelligence. i'm exploring c/c++, javascript, frontend development, cuda optimization, model inference, and ai agents.

i'm especially interested in free and open-source software, privacy, user control, and making powerful ai more accessible without depending entirely on proprietary services or expensive hardware.