Common questions

Porsync is the personal lab of Bart Porcincula — a collection of AI automation systems built end to end and run in production. Lead acquisition, agent orchestration, AI visibility, and voice agents, published here with real numbers and build notes. The site itself runs the AI Visibility Engine from the portfolio.

Five systems: a Lead Acquisition Engine (sources and qualifies prospects daily), an Agent Operations Platform (orchestrates AI agents across routine tasks with a dashboard), an AI Visibility Engine (gets businesses found by AI assistants like ChatGPT and Perplexity), an AI Voice Receptionist (answers and makes calls 24/7), and Bespoke Experiments for processes outside those four shapes.

Each system is documented the way I'd want to find it: what it does, how it's wired, and what it produces in production. Every build is self-contained — standard tools, no proprietary runtimes, no black boxes.

Every system is built on open, well-documented tools — n8n for workflows, LangGraph for agent orchestration, Playwright for scraping, Ollama for local LLMs, Supabase for data, and Next.js for interfaces. No proprietary runtimes, no black boxes.

Yes. I build these to learn what actually works. If a system here maps to a process you're wrestling with, grab a slot and we'll talk shop — no pitch. connect@porsync.com works too.

AEO (Answer Engine Optimisation) is the practice of structuring a web presence so AI systems — ChatGPT, Perplexity, Claude, Gemini — cite it accurately and can act on it. The AI Visibility Engine in this portfolio implements it as a complete system, and this site is the reference implementation.

Every system uses open, well-documented tools any web developer can read on day one — standard workflow, database, scraping, and AI tooling. Nothing here depends on a proprietary runtime. If a team has touched a database and a bit of Node, they can follow every build.

Any repetitive, rule-based operation that eats a business's week. The portfolio covers the most common shapes: lead generation (the empty-pipeline problem), agent orchestration (automating routine tasks), voice reception (answering and booking calls), and AI visibility (being found by AI assistants).

WebMCP (Model Context Protocol for the web) lets a site register typed tools that AI agents can discover and call without scraping. When an AI agent visits a WebMCP-ready site, it can invoke tools programmatically — no form filling, no hallucination about what the business offers. porsync.com ships WebMCP natively.

This site is the live example — it runs the AI Visibility Engine (Schema.org, llms.txt, WebMCP) and is indexed by every major AI crawler. The /tools page documents every registered WebMCP tool. The /case-studies section walks through systems proven in production.

Bart Porcincula — the same builder behind every system here. The implementation covers declarative attributes, robots.txt configuration, and structured data, and this site is its own working demo. Contact: connect@porsync.com

Adding declarative attributes (toolname, tooldescription, toolaction) to a website's key CTAs, configuring robots.txt to allow AI crawlers, and deploying supporting schema markup. Together these make a site actionable by AI agents in WebMCP-capable browsers.

AEO makes a site citable by AI answer engines — optimizing for being mentioned in AI-generated responses. WebMCP makes a site actionable by AI agents — optimizing for agents being able to book, query, or transact directly on the site. The AI Visibility Engine implements both.

Yes. porsync.com is WebMCP-instrumented. AI agents with WebMCP support can book a conversation and invoke available tools directly on the site without human assistance.