LLM Agent Systems
I design workflows where agents operate through missions, providers, memory, safety gates, documentation and human-in-the-loop control.
Agent orchestrationCLIMulti-provider
Available for US-based remote work and AI product contracts
AI Product Engineer
building agentic systems, multimodal apps and developer tools.
I turn ambiguous product ideas into working AI-first prototypes, internal tools, LLM agent workflows and documented engineering handoffs using TypeScript, React, Node.js, Python and Google Gemini.
AI-Born Engineer, not merely AI-assisted. I learned software inside the AI era and use agents as an extension of research, architecture, implementation, testing and documentation.
Professional direction
From AI-native experimentation to business-ready delivery.
The updated portfolio reframes the Rabelus Lab ecosystem as concrete professional capability: agent systems, AI product prototyping, multimodal interfaces, developer tools and technical documentation that can support small jobs, medium builds and larger product initiatives.
AI Product Prototyping
I transform product theses into interfaces, proof-of-concepts, roadmaps, PRDs and technical handoffs that teams can evaluate and extend.
MVPReact/ViteHandoff
Multimodal AI Apps
I build browser-first AI experiences involving chat, voice, image, music, local persistence and user-facing interaction flows.
GeminiVoiceLocal-first
Selected work
Projects that prove the operating range.
Instead of showing every experiment, this section highlights the projects that best translate the GitHub ecosystem into evidence a US company can evaluate quickly.
Tessy / Argenta Fênix
AI-assisted developer workspace and orchestration superproject.
Demonstrates developer tooling, AI-assisted engineering workflows, synchronization contracts, product planning and a flagship direction for agent-powered workspaces.
Developer toolsAgentsArchitectureDocs
Open repository
Inception Framework
TypeScript-native runtime for autonomous AI agents.
Provides the strongest evidence for agentic architecture: missions, providers, CLI patterns, operational control and safety-oriented execution design.
TypeScriptAgent runtimeCLISafety
Open repository
Talia
Browser-first multimodal AI studio powered by Gemini.
Shows product UX around multimodal AI: chat, voice, image, music, local-first storage and a studio-like interface for creative and operational workflows.
GeminiMultimodalReactIndexedDB
Open repository
SOL-Jobs
Solana-first freelance escrow protocol and hackathon product hub.
Documents a product thesis, hackathon scope, PWA-first roadmap and escrow-centered MVP direction for a digital work marketplace on Solana.
Product strategyWeb3SolanaPRD
Open repository
Rabelus Lab Public
Public architecture and manifesto for the AI-first ecosystem.
Consolidates the ecosystem, its verticals, operating philosophy, multilingual documentation and the broader product architecture behind the lab.
EcosystemArchitecturei18nDocumentation
Open repository
DiData
AI learning architect for adaptive knowledge products.
Represents the education vertical: transforming dense knowledge into structured learning experiences, adaptive guidance and AI-mediated study flows.
EdTechKnowledge systemsAI tutoringUX
Open repository
Evidence matrix
Capabilities mapped to concrete repositories.
This replaces subjective skill percentages with verifiable evidence: what was built, which projects prove it and how that translates into a business or engineering use case.
| Capability | Evidence | Representative projects | Business translation |
|---|---|---|---|
| LLM agent systems | Mission workflows, providers, orchestration concepts, CLI and governance patterns. | Inception, Argenta Holistic Control | Internal automation, agentic workflows, AI operations and controlled autonomy. |
| AI product prototyping | Product theses, MVP scopes, PRDs, roadmaps, demos and handoff documents. | SOL-Jobs, Talia, DiData | Rapid validation of product ideas before full engineering investment. |
| Developer tooling | Workspace concepts, sync contracts, terminal/IDE direction and agent-assisted engineering flows. | Tessy / Argenta | Engineering productivity, internal tools, AI-assisted code operations and workflow acceleration. |
| Multimodal AI | Browser-first interfaces involving chat, voice, image, music and local persistence. | Talia | Customer-facing AI experiences and creative/productivity studios. |
| Technical documentation | READMEs, decision records, PRDs, roadmaps, architecture documents and agent-readable handoffs. | Rabelus Lab, Rabelus Lab Public, SOL-Jobs | Cleaner collaboration between founders, engineers, AI agents and stakeholders. |
| Applied vertical thinking | Education, Web3 work marketplace, healthcare concepts, knowledge systems and AI-native productivity. | DiData, SOL-Jobs, DoctorVerso concepts, Rabelus Lab | Adapting AI systems to real vertical problems rather than generic demos. |
Metrics note: the current numbers are based on a June 2026 audit of accessible GitHub repositories. The public GitHub build phase is documented from late 2025 onward, while the broader AI/software immersion narrative starts in January 2024.
Engagement models
How I can support small, medium and large initiatives.
The portfolio is now oriented toward action: it explains what a company can hire me to build, audit, prototype or document.
Small jobs
AI workflow audits, prompt systems, landing/proof-of-concept pages, prototype repair, README/documentation upgrades and focused automation.
1–2 weeksClear scope
Medium builds
Internal AI tools, dashboards, multimodal prototypes, LLM-integrated React apps and agent workflows with technical handoff.
3–6 weeksDemo + docs
Large initiatives
AI-first product ecosystems, developer tooling, end-to-end MVPs, agent platforms and operational documentation for ongoing development.
2–4 monthsArchitecture + roadmap
GitHub interpretation
High-intensity solo development sprints.
The GitHub record shows concentrated bursts of construction, especially around March and April 2026. This is best represented as an intense AI-first build phase rather than a claim of long daily continuity.
Jan 2024
Beginning of declared AI/software immersion and AI-native learning path.
Sep 2025
Observed GitHub account creation window for the current public identity.
Late 2025–2026
Public GitHub build phase documented through repositories, commits and project artifacts.
2026
Rabelus Lab becomes the umbrella for agentic systems, multimodal apps, developer tools and vertical product experiments.
Technical stack
Practical stack for AI-first products.
The portfolio intentionally avoids subjective percentage-based skill claims. Instead, it highlights technologies repeatedly found across the repositories and connected to delivered artifacts.
TypeScriptReactViteNode.jsPythonGoogle GeminiLLM APIsIndexedDBCLI/TUIGitHub ActionsTechnical writingProduct strategy
Resumo em português
Portfólio bilíngue, direção internacional.
Esta versão prioriza o inglês para empresas americanas, mas mantém a explicação em português para preservar contexto, identidade e clareza estratégica.
Posicionamento
Sou um construtor AI-first focado em transformar ideias ambíguas em protótipos funcionais, ferramentas internas, workflows com agentes, documentação técnica e experiências multimodais. A assinatura “AI-Born Engineer” permanece como marca, mas a apresentação principal passa a ser profissional e verificável: AI Product Engineer, LLM Agent Systems Builder e Full-Stack AI Prototyper.
Direção para contratação
O objetivo é mostrar para empresas americanas que o Rabelus Lab não é apenas uma visão filosófica, mas uma prova concreta de capacidade: arquitetar sistemas com IA, prototipar produtos, documentar decisões, criar automações e operar com agentes em projetos pequenos, médios e grandes.
Contact
Let’s build a focused AI-first system.
I am open to remote work, contract projects and collaboration with US-based teams looking for AI product prototypes, internal tools, agent workflows or technical documentation.