A selection of AI builds across voice, lead management, paid acquisition, autonomy, and the web — each built for the same reason: to let a team do what it does best while the system handles the rest. Client identities are kept private; the work speaks for itself.
Whatever you do, do it all for the glory of God — 1 Corinthians 10:31When a hot lead came in, every minute of delay cost a conversion. I built an AI phone agent that fires an outbound call within seconds of a lead arriving — it answers questions from a knowledge base, books a mandatory backup appointment, and live-transfers straight to the owner with zero friction.
The hard part wasn't the talking — it was the reliability: natural turn-taking over noisy phone lines, never cutting off mid-sentence, and guaranteeing the call always ends in a booked slot or a clean handoff.
Outcome: leads reached the moment intent is highest, with a booked appointment as the safety net every single time.
Not every lead is worth a call. For colder, top-of-funnel leads I built a conversational SMS agent that opens the conversation by text, qualifies interest, answers the first round of questions, and only escalates to a live call once the lead is warm — protecting the team's time and keeping the phone for people ready to talk.
It's the front door to the same funnel as the phone agent: text to vet, call to close.
Outcome: a self-running qualification layer that hands sales only the leads worth a human conversation.
A local business was paying for clicks that went nowhere. I rebuilt the funnel end-to-end: a geo-targeted Meta campaign optimized for lead conversions, pointed at the correct conversion form, with a tracking pixel firing a Lead event so the platform optimizes toward booked business — not vanity metrics.
I found and fixed the silent killer too: live ads quietly pointing at the wrong form in the wrong account. Spend was finally chasing the right outcome.
Outcome: a correctly-instrumented acquisition engine where every dollar is measured against real leads.
A study in safe, accountable autonomy. I built an agent that wakes up on a schedule with no memory, reads its own operating manual and journals, does live research, makes decisions inside hard-coded risk guardrails it cannot override, writes what it learned back to memory, and emails a daily report.
Circuit breakers, position limits, and a hard floor sit outside the model's control — the agent is aggressive when the edge is real and does nothing when it isn't. The real product here is trustworthy autonomy.
Outcome: an agent that operates unattended for days while staying inside guardrails — and reports on itself every step.
I design and build high-craft websites with AI in the loop — multi-page, on-brand, responsive, and built to capture leads rather than just look nice. Real brand assets in, a tight design system, and an iterate-until-it's-right screenshot loop.
The site you're reading right now is one of them.
Outcome: sites that look premium and are wired to turn visitors into conversations.
Let's find it and close it.
Start a conversation →