AI systems builder | agent orchestration | local-first runtimes | practical automation
English · Português · Español · 日本語 · 한국어 · 简体中文 · Italiano · Français · Русский · Polski · हिन्दी · العربية · עברית · Bahasa Melayu · Bahasa Indonesia
I build AI systems, agent tooling, local runtimes, and automation layers that turn ideas into execution.
My goal is to contribute to humanity through practical AI work: tools that help people think better, build faster, automate responsibly, and run complex work with more clarity. I want this body of work to be recognized for its impact, including by a great company that values ambitious engineering applied to real human progress.
On-device Universal State Runtime for LLMs on Apple Silicon using MLX, Metal, ANE-oriented architecture, C++17/20, and local-first inference paths. |
Execution CLI for turning one-line software tasks into mapped changes, diffs, tests, validation evidence, and reviewable implementation loops. |
Stack-neutral project mapper for agentic development: shared `.simplicio` artifacts, project context, starter structure, and repeatable AI coding workflows. |
Capability-addressing pattern for agent systems: yool, tuple-space, HAMT catalogs, guardrails, receipts, and reference kernels for large-scale orchestration. |
Hermes Agent performance branch with runtime benchmarks, visual comparisons, low-latency paths, and safe hot-path optimization research. |
Multi-agent sprint delivery workflow for reading sprint systems, mapping repo architecture, dispatching implementation work, and verifying delivery context. |
Skill bundle for Claude, Codex, Hermes, OpenClaw, Cursor, Windsurf, and generic agents covering media generation and LLM workflows. |
One installer and CLI for WaveSpeedAI models across agentskills.io-compatible hosts and multi-agent environments. |
Provider-agnostic AI marketing CLI: brief, script, creative, caption, compliance, publish, metrics, ads, and environment-driven model routing. |
Standalone skill repo for optimizing X posts with source-grounded For You algorithm references, tools, and tests. |
Windows edition of the local inference runtime across NVIDIA, AMD, Intel, NPU, CUDA, DirectML, Vulkan, AVX, and Windows ML targets. |
Local-first Flask panel that turns x.com bookmarks into an actionable queue for AI-assisted triage and execution. PHP/MySQL condominium management system with API foundations for mobile app workflows. Open Finance and Brazilian banking CLI experiments across BB, BTG, Inter, Matera, PagBank, PicPay, and Open Finance BR. Additional work across C#, PHP, TypeScript, Rust, Python, C++, local tools, SaaS prototypes, and AI infrastructure experiments. |
| Project | Stack | Focus |
|---|---|---|
| simplicio-dev-cli | Python | Verified implementation loop for agentic software tasks |
| simplicio-sprint | Python | Multi-agent sprint delivery and repo-context verification |
| ds4-simplicio-apple-v6 | C++ | Apple Silicon local LLM runtime |
| ds4-simplicio-windows-v6 | C++ | Windows local LLM runtime |
| simplicio-marketing-engine | TypeScript | Provider-agnostic AI marketing automation |
| sistema-sindico | PHP | Condominium management system and API foundation |
| brazil-bank-open-finance-br-cli | Python | Open Finance BR CLI research |
| brazil-bank-bb-cli, btg, inter, matera, pagbank, picpay | JavaScript | Brazilian banking CLI experiments |
| phoenix-x-algorithm | Python | X algorithm research and local experimentation |
| bookmarks-oss | OSS workflow | Bookmark-driven knowledge and triage experiments |
- I care about AI that works in real operating environments, not only demos.
- I build systems that preserve context, map repositories, coordinate agents, and automate delivery.
- I am looking for recognition through impact: open work, serious engineering, and opportunities with companies capable of scaling that work.
- The direction is simple: use AI to create more leverage for people, teams, and the future of software.
Special thanks to my mentor Jesse Daniel Brown, PhD, a California, USA native and author of 100+ scientific articles. His humanitarian and educational perspective on programming, AI, and scientific work has helped reinforce my own mission: build practical AI systems that advance human capability.
| Rank | Project | Stars | Forks | Focus |
|---|---|---|---|---|
| 1 | ds4-simplicio-apple-v6 |  |
 |
Apple Silicon local LLM runtime |
| 2 | hermes-turbo-agent |  |
 |
Hermes Agent performance research |
| 3 | simplicio-dev-cli |  |
 |
Agentic task execution CLI |
| 4 | simplicio-mapper |  |
 |
Project mapping for LLMs |
| 5 | simplicio-prompt |  |
 |
Agent capability-addressing pattern |
| 6 | PiAPI-Skills |  |
 |
AI media generation skills |
| 7 | simplicio-marketing-engine |  |
 |
AI marketing automation CLI |
| 8 | simplicio-sprint |  |
 |
Multi-agent sprint delivery |
| 9 | x-bookmarks-panel |  |
 |
Local-first X bookmarks workflow |
| 10 | WaveSpeedAI-Skills |  |
 |
WaveSpeedAI model skills |
Ranking updated from the public GitHub API on 2026-06-01, excluding forks and the profile repository. Star and fork badges update automatically from GitHub. WakaTime remains removed because it depends on external account/data visibility and may render empty or broken in a public GitHub profile. Clone counts and release download counts require GitHub Traffic API authentication and cannot be embedded directly in a public README.
- GitHub: @wesleysimplicio
- Repositories: github.com/wesleysimplicio?tab=repositories
- X: @wesleysimplic
- Instagram: @wesleysimplicio
- LinkedIn: wesleysimplicio
- YouTube: @wesleysimplicio