FlowGenixAI
Selected work

Systems we've shipped.

A look at what we've designed and built over the last two years. Apps, dashboards, automations, and sites, end to end.

AI Operations Dashboard 1600×1000
01Flagship build · Operations

AI Operations Dashboardthe dashboard we run FlowGenixAI on, and deploy for clients

The problem

CEOs and owners of small and mid-sized businesses don't run on one inbox and one calendar. They run on several: multiple email accounts, a stack of calendars, news they can't afford to fall behind on, deadlines, decisions, and a dozen other things scattered across tools that were never built to talk to each other. There's no single place to stand and see it all, so the important work hides behind the noise and the best hours of the day go to hunting for it.

What we built

A private operations dashboard that puts the whole operating day on one screen: every inbox and calendar across every provider, the news that matters, priorities, decisions, and anything slipping, all sorted by what needs the owner first. Custom connectors tie it together past the limits of the native integrations, and the AI triages, drafts, captures voice notes into to-dos, reviews newsletters, and sorts research. It's the same build FlowGenixAI runs internally and deploys, private and branded, for clients.

Stack
Custom connectorsFastAPIClaudeSelf-hosted
View case study →
Almanac Leaf 1600×1000
02Flagship build · Consumer SaaS

Almanac Leafrebuilt and rebranded from Legacy Loop

The problem

The stories a parent most wants to pass on are the ones that slip away, scattered across phones, never written down. Generic journaling apps can't carry that weight, can't hand a message to a child years later, and can't safely let AI help in a domain where a careless tool does real harm.

What we built

A full, live SaaS across web and mobile. Parents journal by voice (recorded, transcribed, summarized), text, or photo, share entries through a private child portal or seal them as time capsules that unlock by date or age, and a governed AI reflects back in one of several mentor voices, always inside a trauma-aware safety layer. Underneath: per-user isolation, private media, a plan and admin model, and a security-audited backend.

Stack
Next.jsSupabaseOpenAIExpo
View case study →
EverSage 1600×1000
03Flagship build · Voice AI

EverSagea voice-first iOS assistant for busy owners

The problem

Running your day by hand, tapping through a calendar, three inboxes, a task list, and your messages, eats the exact hours that should go to real work. And the voice assistants that come on your phone are shallow: they'll set a timer, but they can't triage your inbox, draft a reply that sounds like you, or move a meeting and tell the other person.

What we built

A native iOS assistant you hold a real conversation with. Tap the orb, speak, and it acts: it turns a spoken note into tasks, sets true native alarms and reminders, pulls your inboxes into one triaged list and drafts the replies, opens a text prefilled and waits for your tap, reads your day, and proposes new meeting times. Every action is a typed tool, so what the voice does today a background assistant can do later. Multi-user from day one, and built so your logins and data never sit on the phone.

Stack
React NativeSupabaseOpenAI RealtimeClaude
View case study →
Restate 1600×1000
04Personal build · AI study platform

Restatebuilt for a returning student

The problem

A career changer heading back for an intense graduate program, seven weeks out, with cold study habits and no tool that fit. Generic study apps don't know the calendar, don't teach the workflow, and their AI invents facts, which is a dealbreaker when one made-up citation can sink you.

What we built

A complete, installable study platform, live before prep even started: a dated 7-week ramp, flashcards, quizzes, case briefs, cold-call drills, outlines, a planner, and a deadline tracker on the school's real calendar. The AI is genuinely useful and deliberately fenced in, it works only from the student's own notes, so it can't invent law.

Stack
Next.jsSupabaseOpenAIPWA
View case study →
AI Voice Agents 1600×1000
05Voice AI

AI Voice Agents

The problem

Every missed call is a missed job, but staffing a front desk that answers around the clock isn't realistic for most businesses.

What we built

Retell-powered receptionists and intake agents that answer, qualify, and book across phone and web, day or night, never a ring-out, handing every booking straight to the calendar.

Stack
Retelln8n
Web & app development 1600×1000
06Build

Web & app development

The problem

A great product still needs a fast, credible site and app to turn interest into customers, and most teams don't have the time to design and ship one end to end.

What we built

Full Next.js websites and web apps, designed and shipped end to end on a Vercel and Supabase stack, fast, clean, and built to convert.

Stack
Next.jsVercelSupabase
Workflow Automation 1600×1000
07Automation

Workflow Automation

The problem

Teams lose hours every week to manual handoffs, copying data between CRMs, sheets, and inboxes, and the work that falls through the cracks is invisible until it's too late.

What we built

Custom multi-step workflows that connect CRMs, sheets, voice agents, and AI to move work automatically between tools, with clean reporting on every run, so nothing is invisible.

Stack
n8nAPIsSupabase

Have something like this in mind?

Book a strategy call