A seven-location US restaurant group had a problem it couldn’t see on the menu. It was sending about $60,000 a month through third-party delivery apps, and after fees and promotions the real commission was close to 28%. With a net margin of only 3 to 4%, most of those delivery orders earned the group almost nothing. The worst part was that the majority came from regulars who already loved the food and would have ordered direct if it had been easy. We built Ember, a branded iOS, Android, and mobile web ordering app, and rolled it out across all seven locations in four months. The goal was never to replace DoorDash, Uber Eats, or Grubhub. They still do a great job introducing the restaurants to new customers. Ember simply gives existing customers a faster, more rewarding place to reorder, on a channel the restaurant owns. What we built: Native iOS and Android apps, plus a mobile web ordering page so customers can order without installing anything Full POS and kitchen display integration across all seven locations, so every order reaches the line with no manual re-entry A loyalty program, push notifications, and smart checkout upsells An opted-in customer list the group owns and can market to for free A three-phase rollout, starting with a two-location pilot to prove the numbers before going to all seven Results: 35% of repeat delivery orders moved from marketplaces to direct ordering About $63,000 per year in commission recovered 22% higher average order value on direct orders Around 20% less order-handling labor during peak service
StackNovaTech
We Build Software That Scales Your Business
StackNovaTech Portfolios
Group trips rarely fall apart over the destination or the budget. They fall apart over coordination and money. Most groups end up juggling three separate apps, one for the plan, one for decisions, and one for splitting costs, and none of them talk to each other. Flock is a group travel app built around one idea: a group’s decisions and a group’s money are the same thing. The person who suggested the villa, the people who voted for it, and the person whose card got charged should all be looking at one screen. Flock puts the itinerary, group chat, documents, voting, and live budget into a single shared trip. The research behind it: 54% of US adults took a group trip in the past five years (CIT Bank / Harris Poll, May 2026) 45% of group travelers have run into financial conflict or discomfort on a trip 82% would pay more than their fair share to avoid a money argument What we built: One shared trip with the plan, chat, confirmations, and budget in the same place Expenses tied to the decision they came from, so costs are split across the people who committed instead of just divided by heads Live balances that show who is carrying more than their share while the trip is still happening Lightweight voting on dates, destinations, and activities, plus assigned tasks with owners and due dates, so one person doesn’t end up as the unpaid planner Simple onboarding and expense entry, designed for the member who joined reluctantly Native iOS and Android apps on a shared backend, with real-time sync so everyone sees the same trip the moment something changes
An online travel agency selling flights, hotels, and holiday packages through its mobile app was losing bookings at the two moments that decide loyalty: the last step of checkout, and the hour a trip falls apart. The app could sell a trip, but it could only show information, never act on it. Travelers filled out long forms to book, and when a flight was cancelled they were left in the airport with an alert and a support queue. We built Voyagent, an agentic AI travel assistant that plans, books, and rebooks on its own. The key design decision was where to draw the line on autonomy. The agents do the work, and the traveler approves every moment where money or a booking is on the line. That keeps trust intact while removing the effort. What we built: Natural language trip planning, so travelers describe what they want and get a short set of bookable options instead of a form One-tap booking on stored preferences and passenger data, with a single approval step to confirm the final price A monitoring agent that watches every flight segment and presents ready-to-confirm rebooking options the moment a disruption hits A preference profile that learns from each trip, so every round of recommendations is sharper than the last Four agents (planning, booking, monitoring and rebooking, and preference memory), each with a self-check loop that verifies options against the traveler’s constraints before showing them Guardrails including human approval on every booking, hard budget limits, and a full log of every agent action for support to audit A React Native app connected to live flight and hotel inventory through the client’s supplier APIs Results: 31% drop in booking abandonment Disruptions resolved in under 60 seconds, from alert to confirmed rebooking 22% lift in repeat bookings 40% fewer disruption-related support tickets Delivered by a team of five: one product lead, two engineers, one AI engineer, and one designer.
Jammin Car Wash is a family owned operator with two express wash sites in Lutz and Wesley Chapel, Florida, and a third under construction on SR 56. The business doesn’t really sell car washes. It sells the Daily Wash Club, a recurring membership at $25 to $50 a month. That meant the website had a tricky job: convince visitors to choose a subscription, even though the site itself can’t take a payment or create an account. Signup, login, and plan management all live on an external platform, mywashaccount.com. A first-time visitor also had to work out, on their own, whether a $14 single wash or a monthly plan made more sense, and five different goals (retail washes, memberships, fleet services, fundraising, and recruiting) were all competing for the same homepage. We redesigned the site so the decision happens on the page, before the handoff. Visitors compare all four wash tiers and twelve price points in one view, pick a plan, and arrive at the membership platform ready to sign up instead of browse. What we built: A restructured homepage where the four-tier wash comparison, not the hero, drives the conversion Member login and plan management in the top utility navigation, so existing members skip the sales flow completely Separate pages for family and annual plans, fleet services, the fundraising program, and a combined contact and FAQ page, with recruiting moved to an off-page application form A WordPress build on a custom child theme, with image-heavy components compressed against a set performance budget Signup and login routing to the external membership platform A location block with address, phone, hours, embedded map, and directions for all three sites, including the one still under construction Results: Desktop loads in 1.5 seconds on a page built almost entirely from image components Four wash tiers across twelve price points in a single comparison view Three locations on one page, each with its own address, hours, map, and directions Delivered by a team of five: one team lead, two developers, and two designers.
Rock Solid Health is a single-location wellness clinic in Round Rock, Texas, offering colon hydrotherapy, colonoscopy preparation, ionic foot bath, and infrared sauna sessions. Most local clinics grow through referrals, but this category has none. People don’t bring up colon hydrotherapy over lunch, and they rarely call a clinic to ask what a first session involves. They search privately, read privately, and either book or leave. That meant the website’s content had to do the work a friend’s recommendation normally would. We built the site around that constraint. Every question a visitor might be too uncomfortable to ask out loud is answered on the page, before the booking step, so they arrive at the booking page already reassured. The clinic also serves a second audience with very different needs: colonoscopy preparation clients working to a fixed medical date. They get their own service track and call to action, so they never have to wade through wellness browsing content. What we built: A service structure covering colon hydrotherapy, colonoscopy preparation, ionic foot bath, and infrared sauna, each with its own page and booking path A WordPress build on a custom child theme, with a video hero and a pricing page with plan options An educational content set covering first sessions, open systems, colonoscopy preparation, and the questions clients arrive with Online booking against real availability, including the clinic’s last-appointment cutoff, plus enquiry routing, so scheduling happens through the site instead of the phone Factual service descriptions that explain what happens in a session, with claim language left under the clinic’s control Results: In production since 2020 and still running the practice, having absorbed a new service line without a rebuild Four services bookable online against real availability 2.3 second mobile load and 2.0 second desktop load, on a page with a video hero Delivered in 2020 by a team of five: one team lead, two developers, and two designers.