v1.0.0
    QWRM

    QWRM Case Study

    QWRM was created to make it easier for people-especially busy New Yorkers-to actually meet up with friends instead of letting plans stall in endless group chats. Missed replies, vague maybes, and last-minute changes often derail great ideas; QWRM solves this by turning real-world invitations into simple, commitment-based actions.

    Designed as a social planning tool-not a social network-QWRM helps users quickly find friends or acquaintances for concerts, movies, sports, picnics, and more. The platform is intentionally private: no public profiles, no algorithmic suggestions, no invites from strangers.

    QWRM showcase

    Challenge

    One notable hiccup was converting the Bolt codebase from React.js to Expo - something that turned out to be harder than expected. Another was that the intended name of the app was already taken on the App Store, which involved coming up with another name that captures the proposition. QWRM, short for Quorum, seems to capture the value pretty well.

    QWRM detail

    Development & Implementation

    After creating wireframes in Figma, our team got to work setting up AWS and Facebook, Google Cloud and Sendgrid accounts and using AI to generate code from the wire frames. We started by creating the app from scratch using bolt.new, and followed that up with Augment to streamline development and automate many of the heavy lifting tasks.

    What worked especially well was splitting our environment into two Visual Studio Code instances — one to keep the agent context live, and another for direct, chat-based edits. This setup let us move quickly while staying focused and organized.

    In a second iteration, the mobile app received a major round of polish, with a refreshed visual design, improved event listings, and streamlined account creation. Key behaviors were tightened around permissions, notifications, and editing tools, making the app more intuitive and secure. Several infrastructure fixes were also completed, including email reliability, domain updates, and Mixpanel integration for analytics.

    QWRM feature

    Results

    Prior to AI, a traditional app like this would've required about 2 weeks of design and at least 6 weeks of development. We completed it in just 117 hours. That's the power of AI-driven development.

    Location:

    NYC

    Tech Stack:

    Frontend: React Native, Expo

    Server Host: AWS

    Third Parties: Sendgrid, Mixpanel

    Features:

    • Authentication & Login
    • Account creation
    • People & Contacts
    • Event Management
    • Invitations
    • Email and Push Notifications
    • Payments

    Achievements:

    • [Achievement 1 - e.g., Reduced quality issues by X%]
    • [Achievement 2 - e.g., Improved resource utilization by Y%]
    • [Achievement 3 - e.g., Served Z enterprise clients]
    Beth Sullivan
    Beth Sullivan
    Co-Founder
    "Before meeting Oscar, all we had was what we felt was a great idea and a set of killer wireframes. Oscar and his team at Blood and Treasure were able to take those wireframes and turn that idea into a fully functional app, available for both iPhone and Android. Aside from writing the code for each native version of the app, Oscar has been instrumental in navigating the labyrinth of requirements and approvals necessary to get the app out of beta and into the app stores. When we decided to undertake a total rebrand and redesign, Oscar and his team were able to pivot quickly and get all of our changes in place without missing a beat. And the collaboration hasn't ended now that the app has hit the stores. As we've collected feedback from users, we've continued to rely on Blood and Treasure to implement new features and come up with creative solutions to things we might have missed. There have been a lot of "aha" moments on our recap calls. Billing from Blood and Treasure has always been fair and without surprises. We truly feel that Oscar cares more about creating a great product to our specifications than he does about racking up unnecessary billable hours."

    Still have questions?

    Talk to Devin, our AI assistant, for instant answers about our services.