AI chatbots, voice agents and lead-routing pipelines running for real businesses. A live SaaS product of my own, built with a multi-agent Claude Code team I run daily. And 6 years of production engineering behind all of it, including banking apps on secured backend APIs.
I'm John Dominic, an AI Automation Engineer. Most of my work is building intelligent automation for businesses: AI chatbots, voice agents, lead-routing pipelines, and full workflow automation, running in production for real clients.
I also build and run LiquidityHQ, my own live SaaS product for crypto traders. I built it with a multi-agent Claude Code team I run every day: developer, QA and PM/DevOps agents pick up GitHub issues, write and test the code, and promote it from dev to staging to production. I review the work and approve anything that touches the live database or ships visually, so I know what it takes to ship and operate something real, not just prototype it.
Underneath all of it is 6 years of production software engineering, including banking apps connected to secured backend banking APIs, where I handled sensitive customer data, and Android development in Kotlin. That is where the reliability habits come from, and they are the reason these systems hold up: my ingestion service rejects duplicate work with idempotency keys instead of double-charging a customer, and a circuit breaker stops calling a failing downstream step rather than hammering it with retries. Built to degrade in a controlled way, not demoed once and abandoned.
Two toolsets, one goal: ship reliable systems fast.
Automation & AI
Web, Mobile & Backend
Production systems for real businesses, prototypes built to production standards, and a SaaS product of my own. Each one targets a specific manual process or failure point.
Retail traders drowning in scattered market data across a dozen tabs, acting too late.
A live SaaS product I built and run. Scores squeeze setups and whale activity across 50 coins in real time, with Grok reading 35 signals to give a direct trade bias. Built with my multi-agent Claude Code team.
Leads slipping through the cracks across calls, forms, and follow-ups, with no single system catching them.
5 systems unified into one B2B automation machine: Dead Lead Reactivation, a missed-call voice receptionist, PropConnect's voice agent, Smart Lead Routing, and an AI Appointment Setter.
Security alarm systems that drown operators in false positives, or silently drop real threats when events arrive in bursts.
A real-time alarm monitoring prototype I built end to end: WebSocket ingestion under bursty load, a computer vision camera worker, AI triage with fallbacks and a spend cap, a live dashboard, and pattern-based escalation.
Bookkeeping automation that breaks quietly: duplicate charges slip through and one outage takes the whole run down.
An ingestion service with idempotency-key protection and a circuit breaker that stops calling a failing downstream step instead of retrying blindly.
Branches fielding inquiries by hand, with replies taking hours and no consistent coverage.
A router agent classifies each incoming conversation and dispatches it to a specialist sub-agent, answering instantly across every branch, 24/7.
Android apps built end to end, from idea to release, in Kotlin.
Cyclists stranded with no fast way to call for roadside help.
A stranded rider taps once and the nearest available helper gets their exact GPS location, no directions to explain. Both positions stay live on the map as help gets closer.
Missing your stop because there is no heads-up before it arrives.
Set your stop and relax. The app watches the route in the background and wakes you before you arrive, even with the screen off, so a long ride no longer ends in an overshoot.
Each group below came out of a system that shipped and is running. The note under each one is the architecture it was built on.
Backed by live asynchronous webhook triggers, contact syncing loops, and native CRM booking automations built inside GoHighLevel.
Engineered on the same retrieval-augmented generation (RAG) stack running in my production engines: text embedding, semantic search, and pgvector routing.
Built on background processing, real-time database syncing, and transactional error-proofing developed across my Android apps and FastAPI backends.
Hiring for AI automation, or want any of the systems above walked through? Tell me what you're building.
Prefer a person? Book a call or email me.
