I build software at the intersection of LLM systems, developer tooling, compilers/runtimes, and scientific computing.
I care about tools that make complex workflows executable, inspectable, and benchmarkable.
I now work in full human-AI symbiosis: whether I am job hunting, writing papers, or replying to emails, I stay synchronized with my private assistant. If my assistant and I cannot reach agreement on a product—whether by the assistant convincing me to revise it or by me convincing the assistant to accept my taste—I will not promote it. Anything that an AI can already identify problems with does not deserve to waste human attention.
- Working on agent skills, MCP servers, model-serving infrastructure, compiler/runtime projects, and evaluation tooling.
- Strongest with Python and Rust; usually shipping CLI tools, notebook tooling, and GPU/HPC-heavy workflows.
- Interested in software engineering, ML systems, and research engineering roles in Europe, the UK, and the US.
|
skills Installable agent skills for Codex and Claude, with marketplace generation and local sync tooling. |
swarmx A lightweight multi-agent orchestration framework with MCP support and an OpenAI-compatible server. |
|
yallm A high-performance API hub that maps between clients and multiple model providers with streaming-first compatibility. |
xcc A modern Python C compiler focused on deterministic behavior, transparent diagnostics, and a clean architecture. |
|
wenyan A Python implementation of the Wenyan programming language with CLI tools, benchmarks, and a self-hosted path. |
JupyMCP An MCP server for rich Jupyter notebook interaction, including cell editing, execution, and notebook resources. |
|
miniprot-rs A Rust reimplementation of miniprot focused on clarity, maintainability, and performance, with room for embedding-augmented workflows. |
cotton2k A Rust rewrite of an established scientific simulation model with Python bindings and batch workflows. |
- Hardening
yallmandswarmxaround provider compatibility, streaming, tool use, and MCP-native workflows. - Building notebook and local-agent tooling through
jupymcp,skills, and related projects. - Working on language/toolchain projects such as
wenyanandxcc. - Exploring benchmark and evaluation workflows, including domain-specific reasoning tasks.
- Resume: tcztzy/resume
- Notes and references: tcztzy/bubble
- GitHub profile: @tcztzy
- Contact: tcztzy@gmail.com