╭────────────────────────────────────────────────────────╮
│ RAMAILO1 // SYSTEM PROFILE │
│ │
│ STATUS 🟢 ACTIVE & BUILDING │
│ DOMAIN AI · PRODUCT ENGINEERING · DEV INFRASTRUCTURE│
│ MODE SYSTEMS ARCHITECTURE & PRODUCT ENGINEERING │
│ CURRENT SpicySwipe · DriftGuard · Onward90 │
╰────────────────────────────────────────────────────────╯
constramailo1={role: "Product Engineer · AI & Systems Architect",mindset: ["Understand Problem","Design Architecture","Build Clean MVP","Measure Telemetry","Ship & Iterate"],focus: ["Product Engineering & Full-Stack Systems (Next.js / TypeScript / React / Prisma / PostgreSQL)","Applied AI & Agent Tooling (Gemini / Claude / OpenAI / Local Ollama / MCP Protocol)","Developer Infrastructure & Browser Automation"],philosophy: "Don't just write code — engineer products that solve real problems and feel great to use."}; IDEA ──► Understand Problem ──► Design Architecture ──► Build Smallest Useful Version ──► Measure & Scale ──► Ship 🚀
"I care about the intersection of clean engineering architecture and unforgettable product experience."
🔥 SpicySwipe — AI Browser Extension
AI-powered browser extension combining multi-LLM generation, contextual personalization, automated workflows & conversation analytics.
Profile Context ──► Multi-LLM Engine ──► Tone Adaptation ──► Smart Automation ──► Live Telemetry
(Bio Extraction) (Gemini/Claude/GPT) (Adaptive Personas) (One-Click / Auto) (Match Analytics)
- The Problem: Repetitive manual matching and bland conversational outreach lack personalization, contextual awareness, and engagement analytics.
- The Solution: A Manifest V3 browser extension integrating Claude, Gemini, OpenAI, and local Ollama to analyze profile bios, adapt conversational tones, and execute automated workflows with enterprise-grade XSS security.
- Stack:
JavaScript•Manifest V3•Ollama (Local)•OpenAI API•Google Gemini•Claude API•Analytics - Links:GitHub Repository → | Live Demo & Docs →
🛡️ mcp-server-driftguard — AI Agent Governance
A Firewall for Intelligence — MCP server enforcing finite state machines, scope verification & tamper-proof audit trails.
┌────────────────────────────────────────────────────────┐
│ 🛡️ DRIFTGUARD // AI AGENT GOVERNANCE │
│ │
│ AI AGENT (LLM Runtime) │
│ │ │
│ ▼ │
│ ┌──────────────┐ │
│ │ Scope Claim │ │
│ └──────┬───────┘ │
│ ▼ │
│ ┌──────────────┐ │
│ │ Permission │─── ❌ ──► Unauthorized Mutation Blocked│
│ │ Checkpoint │ │
│ └──────┬───────┘ │
│ ▼ │
│ ┌──────────────┐ │
│ │ Audit Trail │─── ✔️ ──► Cryptographic State Log │
│ └──────────────┘ │
└────────────────────────────────────────────────────────┘
- The Problem: Autonomous coding agents drift from scope, execute unintended file modifications, and lack verifiable, auditable safety rails.
- The Solution: An MCP server providing a state-machine policy engine that validates agent claims before filesystem mutations and logs cryptographic audit trails.
- Stack:
TypeScript•Model Context Protocol (MCP)•Node.js•State Machines - Links:GitHub Repository →
🚀 Onward90 — Employee Onboarding SaaS
Day-by-day employee onboarding companion to accelerate ramp-up time and provide HR full visibility.
- The Problem: New hire ramp-up is fragmented, leading to lost productivity and lack of structured visibility for managers and HR teams.
- The Solution: A 90-day interactive onboarding companion with scheduled milestones, manager checkpoints, and full organizational reporting.
- Stack:
Next.js•TypeScript•Prisma ORM•Auth.js•PostgreSQL•TailwindCSS - Links:GitHub Repository → | Live SaaS App →
An end-to-end architecture pattern I use when engineering AI-integrated full-stack applications:
┌──────────────────────────────────────────────┐
│ Responsive Client Interface │
│ Next.js · React · TypeScript · PWA/Ext │
└──────────────────────┬───────────────────────┘
│ (HTTPS / WebSocket / RPC)
▼
┌──────────────────────────────────────────────┐
│ API Gateway & Routing │
│ Authentication · Rate Limiting · RBAC │
└──────────────────────┬───────────────────────┘
│
┌───────────────────────┼───────────────────────┐
▼ ▼ ▼
┌────────────────────────┐ ┌─────────────────┐ ┌───────────────────────┐
│ AI & Agent Runtime │ │ Core REST/tRPC │ │ Background Workers │
│ LLMs · Orchestration │ │ Node.js/Express │ │ Queues & Async Tasks │
└────────────┬───────────┘ └────────┬────────┘ └───────────┬───────────┘
│ │ │
▼ │ │
┌────────────────────────┐ │ │
│ MCP Tool Protocol │ │ │
│ Tools · Policy · Guard │ │ │
└────────────┬───────────┘ │ │
│ │ │
└───────────────────────┼───────────────────────┘
▼
┌──────────────────────────────────────────────┐
│ Data & Storage Layer │
│ PostgreSQL · Prisma · Redis Cache · S3 │
└──────────────────────────────────────────────┘
┌────────────────────────────────────────────────────────────┐
│ 🧠 AGENTIC AI │
│ MCP · Tool Calling · State Machines · Evaluation │
│ │
│ 👁️ SPATIAL & COMPUTER VISION │
│ Vision-Language Models · Spatial AI · Verification │
│ │
│ ⚡ DISTRIBUTED SYSTEMS │
│ Queues · Asynchronous Workers · Event Architecture │
│ │
│ 🌐 WEBGPU & INTERACTIVE WEB │
│ Real-Time Canvas · 3D Interfaces · Micro-Motion │
└────────────────────────────────────────────────────────────┘
Have an interesting product idea, technical challenge, or looking to collaborate?
- 🚀 Building: AI-powered products, developer infrastructure, MCP tooling, or full-stack web platforms.
- 💬 Reach Out:
