Running a family
of five shouldn't
run you.
DeLucaHub is a custom-built automation platform that handles the daily logistics of family life — silently, reliably, and without asking for your attention.
Modern family life generates an impossible amount of signal noise.
Calendars, email, texts, school apps, sports schedules, bills, reminders — none of it talks to each other. Staying on top of your family's week requires constant context-switching between a dozen different tools.
When's the next thing happening at school? Which birthday party have we already committed to, and is the gift handled? How much have we already spent this week? That kind of ambient knowledge lives rent-free in someone's head — and the moment it slips, the cracks start to show.
Five people. Two careers. Three kids — each with their own social calendar, school events, and a steady stream of birthday party invitations. Every single one means a last-minute trip to find a gift. The off-the-shelf tools never quite fit. You're always adapting yourself to the software, not the other way around.
Every consumer tool is designed to maximize engagement, not reduce it. They want you opening the app. Checking the dashboard. Subscribing to the premium tier. The incentives are backwards — your attention is the product.
"The best system is the one that does its job and disappears. Silence is the success metric."
So I built the infrastructure myself.
DeLucaHub runs on a private Linux server with five purpose-built Python workers in Docker. Each worker handles a specific domain of family life — emails, calendars, finances, birthdays, and vehicles — and surfaces the right information only when it matters.
The dashboard, a Next.js web app installable on Android and iOS, is how new information enters the system cleanly: birthday invitations, school events, vehicle maintenance logs, and gift updates. Structured intake, not casual chat.
Structured intake.
Not another inbox.
The dashboard lives on the web and installs natively on iOS and Android. Every button is a structured data entry point — designed to put information into the system cleanly, so the workers can act on it reliably. No free-text, no ambiguity.
DeLucaHub Dashboard · v1.0 · Web · iOS · Android
Five workers. One job each.
Each worker owns a domain and runs independently. When nothing is urgent, they stay quiet. When something matters, they surface it.
Every morning, a structured summary of what the family needs to know — email threads, school updates, action items — stripped of noise. Powered by Gemini Flash for extraction and classification: fast, accurate, and cost-efficient at the volume of a family inbox running daily.
Proactive look-aheads — 2-day reminders, Monday weekly previews, Friday weekend previews. Nobody misses the thing they forgot was on the calendar.
Tracks transactions across all family credit cards and delivers a daily summary of accumulated spending for the current week — measured against the family's custom budget parameters. A weekly recap follows, breaking down the top categories of spend.
When a party invitation enters the system via the dashboard, it lands on the family calendar and triggers a reminder sequence — 14 days out, 7 days out, day-of. When a gift is needed, the worker routes to a curated Amazon shopping list filtered by the child's age and gender.
A structured, searchable single source of truth for the family fleet. Oil changes, tire rotations, repairs — every service record logged through the dashboard and stored permanently in Supabase.
The architecture is modular. Adding a new worker means writing a new Python service. The shared Supabase backend and Telegram control surface are already there.
A day in the life.
What the family actually experiences — not the code, not the infra.
A clean, structured card: today's calendar events, any action items from overnight emails, school notices, and one financial flag if anything needs attention. Read in 90 seconds over coffee.
Every Monday morning, a heads-up on the full week ahead — who has what, when, and whether there are any schedule conflicts worth knowing about.
A birthday party invitation arrives. Open the dashboard, tap "Add Birthday Party Invitation," enter the party details, and choose which child was invited. DeLucaHub already knows the kids' ages, so it can infer the right gift age range without making you enter extra information. The event lands on the family calendar, and reminders are scheduled.
The reminder sequence fires automatically. Not a frantic day-before scramble — a full week out, when there's still time to order something thoughtful.
Service is done. Open the dashboard, log it in 30 seconds. The record lives in Supabase permanently — date, mileage, what was done.
What's happening this weekend — kids' activities, any plans, anything that needs prep. One card, delivered Friday afternoon when there's still time to prepare.
shaped every
decision.
Every design decision in DeLucaHub started with a question: why do we do this the way we do it? Not an assumption that it was broken, but a genuine curiosity about whether technology and AI had reached a point where they could meaningfully support how our family operates.
The goal isn't to remove thinking — it's to take the predictable, recurring tasks off the mental stack so there's more room for the things that actually require judgment.
Not every pain point belongs inside DeLucaHub. Knowing when not to build is as important as knowing how.
A system that only works for the person who built it isn't a household system — it's a personal project. DeLucaHub has to be intuitive enough for both of us to use independently.
Telegram is for receiving — briefings, alerts, and a lightweight conversation with Claire. The dashboard is for sending — structured data entry that feeds the system cleanly.
Every worker runs on a schedule. Every output is predictable. AI is reserved for judgment at the edges — extraction, classification, conversational response — not as the primary control loop.
The best version of DeLucaHub is the one you stop noticing. Not because it stopped working — because it's working so consistently that it just becomes part of how the household runs.
vision to a
system.
The ideas behind DeLucaHub didn't start with a household roadmap. They started with a habit — a constant low-level scan for where things could work better, and what technology might make possible. The household was just the next place to apply that same logic: take the recurring pain points of family life, and ask whether there was a better way to handle them.
In early February 2026, through what was then known as ClawdBot — in the middle of its rapid evolution into Moltbot and then OpenClaw — something shifted. The concept of a personal AI that learns you, runs locally, and operates through the tools your family already uses wasn't theoretical anymore. It was buildable. That realization was the starting gun.
Starting at home rather than at work was a deliberate choice. No enterprise constraints, no IT approval loops, no organizational politics — just real problems, real stakes, and real feedback on whether something actually worked. What followed was genuine trial and error: learning which problems genuinely benefit from an AI agent's judgment, and which are better served by a hardened, deterministic process that simply runs. That distinction became one of the most important design lessons in the whole project.
Two tools in particular changed what was possible on the build side. v0 by Vercel made it possible to describe a design intent conversationally and get back something polished and functional — closing the gap between how something should look and feel and what actually gets built. Claude Code took that same principle to the backend: getting directly into the VPS, understanding context, and managing code more precisely than any agent-based approach had managed before. For backend optimization, it was a meaningful step up. Together, they made the difference between a system that works and one that's actually maintainable.
DeLucaHub today is an MVP — functional, reliable, and already changing how the household operates. But it's also a foundation. The capabilities continue to evolve, the use cases will expand with them, and the system will grow accordingly. What remains constant is the question driving all of it: where AI can genuinely work alongside the way we already live and work, and what becomes possible when someone who thinks in systems finally has the tools to build them.
Built by someone who
always asks if there's
a better way.
DeLucaHub isn't a side project for its own sake. It's the natural extension of how I've approached every system I've ever touched — identify the real pain point, challenge the assumptions behind why it works that way, and then design something cleaner, simpler, and more efficient in its place.
If this sparks something for you — whether it's the systems thinking, the household angle, or just your curiosity about where AI fits into how we'll all work and live — reach out. Always happy to connect and collaborate. Send me a note or find me on LinkedIn.
The better question is whether the process still fits the goal. What friction have people learned to tolerate? What tools exist now that did not exist then? If we were designing the process today, would we still build it this way?
Across my career, that has been the through-line: noticing when everyday processes have become more complicated than they need to be, questioning the assumptions behind them, and looking for a better way. DeLucaHub is that instinct applied at home.