Skip to content
View zaesho's full-sized avatar
馃惐
馃惐
  • Hire Me
  • Your Walls
  • 18:10 (UTC -04:00)
  • X @zaedev

Block or report zaesho

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don鈥檛 include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user鈥檚 behavior. Learn more about reporting abuse.

Report abuse
zaesho/README.md

Hey, I'm Gian! you might know me as Zae.

LinkedIn 路 Broadcast credits 路 Creative work

I鈥檓 a broadcast producer, director, and engineer, a physics student at RIT, and a curious mind who likes digging into as many things as possible!

My work moves between Esports broadcasts, software for the people running them, and experiments in ML. That means most days im at class, or running show. And most nights im enabling my scatterbrain in whatever-the-hell-else I can think of.

From the control room to the codebase

  • Live production: Rainbow Six Siege broadcasts with FRAGADELPHIA, Nerd Street, BLAST, and Crucible Media, including directing the Six Invitational 2024 in S茫o Paulo.
  • Broadcast tools: real-time match data, statistics, graphics workflows, and operator controls that have to make sense when the show is live.
  • Software: desktop tools for coding agents, remote workflows, and the infrastructure that connects them.
  • Research: GPU inference, information retrieval, and predictive coding.
  • Physics: studying at Rochester Institute of Technology, with minors in Mathematics and Quantum Information Science.

A few things I've been working on

Project What it's about
FlashEmbed GPU inference for embedding models, with short-query optimizations, FP8 document indexing, and retrieval-quality checks.
S1Rank An independent study of Jev as a retrieval reranker: benchmarks, raw responses, calibration experiments, and a paper.
PC-ALM Anchor Exploring why predictive-coding inference loses gradient accuracy, and testing forward-anchored damping through theory and experiments.

I also work on forks of MonoCode, T3 Code, Cua, and RustDesk, exploring coding-agent integrations, computer use, and remote access.

Tools I reach for

Code 聽 TypeScript 路 Python 路 Rust 路 SQL 路 React 路 PostgreSQL
Research 聽 PyTorch 路 Triton 路 CUDA 路 Manim
Production 聽 vMix 路 Bitfocus Companion 路 Adobe Creative Cloud 路 Figma


English & Espa帽ol. Equally happy talking about a broadcast rundown, a slow kernel, or why the universe does that.

Popular repositories Loading

  1. phys-isr phys-isr Public archive

    Manim animations explaining special relativity time dilation

    Python

  2. replay-tool replay-tool Public

    Forked from wnc-replay/replay-tool

    Go

  3. rustdesk rustdesk Public

    Forked from rustdesk/rustdesk

    An open-source remote desktop application designed for self-hosting, as an alternative to TeamViewer.

    Rust

  4. zeron zeron Public

    Forked from zeronsh/zeron

    A native control plane for Claude Code, Codex, Cursor, Devin and other coding agents.

    Rust

  5. pc-alm-anchor pc-alm-anchor Public

    Forward-anchored damping makes augmented Lagrangian predictive coding (PC-ALM) converge: theory, experiments, paper

    TeX

  6. decode decode Public

    Forked from zai-org/ZCode

    Z.ai's coding agent harness. Powerful, intelligent, extensible.

    TypeScript