Token Budget Coach: Right AI Model per Task (Claude Code)
by Pedro Alexandre Brito
Picks the right AI model and effort for every prompt, so you save tokens and keep the quality.
Full-stack software engineer specializing in agentic AI ecosystems and developer tooling. I focus on integrating frontier LLMs—such as Claude Code and Gemini—directly into software engineering pipelines to build autonomous workflows, custom plugins, and high-efficiency microservices. My background bridges robust backend engineering (Java/Spring Boot) and interactive modern frontends (TypeScript, React, Vanilla JS) with intelligent agents designed to eliminate repetitive tasks, streamline data operations, and accelerate shipping velocity. What I Build & Offer: Custom Plugins & Extensions: Tailored tooling, MCP servers, and CLI extensions for Claude Code, Gemini, and AI-assisted IDEs. Autonomous Agentic Workflows: End-to-end multi-step automations for code generation, refactoring, auditing, and operations. Full-Stack AI Integrations: Connecting intelligent agents seamlessly to external APIs, webhooks, and cloud databases (Supabase / Neon).