Mobile Application
Patio Masters
A Flutter-based mobile training app that takes new sales hires through a structured onboarding programme and automatically schedules their interview on completion, while giving existing staff AI-generated practice questions grounded in real company context.
Project Snapshot
Platform
iOS & Android
Built With
Flutter
AI Questions
Company-Trained
Onboarding was inconsistent, dependent on whoever was available to train, and gave managers no visibility into who was ready for their next interview. Existing staff had no structured way to keep their product knowledge sharp between jobs.
Ecello built a dual-experience Flutter app with role-based access. New hires progress through locked training segments, pass post-video assessments, and trigger an automated interview booking upon completion. Existing staff access an AI-powered practice section where questions are dynamically generated from the company's own product catalogue, sales scenarios, and objection-handling context.
Every new hire now arrives at their interview with the same baseline preparation. Managers are notified when candidates are ready rather than chasing progress manually. Existing staff have an always-fresh practice tool that requires no additional overhead to maintain.
Overview
About this project
Patio Masters is a mobile training and development application built for a patio installation and sales business. The app serves two distinct groups: new hires who need to complete a structured onboarding programme before their first interview, and existing sales staff who use it on their days off to sharpen their product knowledge through AI-generated practice questions. The entire application was built in Flutter, making it available across both iOS and Android from a single codebase.
The Problem
What we were up against
Sales training in field-based businesses tends to be inconsistent. New hires often shadow someone for a few days, watch some videos if they exist, and then get put into the field with incomplete preparation. The quality of their training depended entirely on who happened to be available, and whether that person was a good trainer or just a busy salesperson trying to meet their own targets. For existing staff, there was no structured way to continue developing between jobs. Product knowledge and sales technique erode without practice, and a salesperson who is not regularly challenged tends to fall back on the same small set of approaches. There was also no way for managers to track any of this.
The Approach
How we solved it
The app has two clearly separated experiences within the same codebase. New hires move through training segments in a fixed order, with each segment locked until the previous one is completed and assessed. Once all segments are done and all assessments passed, the app automatically triggers an interview booking, notifying the manager and scheduling the candidate without any manual follow-up. For existing staff, an AI model prompted with the company's own documentation generates fresh questions every session covering product knowledge, objection handling, pricing, and installation details. No two sessions are identical, and the material never goes stale.
Features
What we built
New hires and existing staff log into the same app but see entirely different experiences. The system serves the appropriate interface automatically based on who the user is.
New hires move through training segments in a fixed order. Each segment is locked until the previous one is completed and assessed, enforcing consistency across every person who goes through the programme.
At the end of each training segment, the user answers a set of questions testing their comprehension. Passing is the condition for unlocking the next segment. Failing prompts a review before they can try again.
When a new hire completes all segments and passes all assessments, the app automatically triggers an interview booking. The manager is notified and a time is arranged without any manual follow-up from either side.
The existing staff section generates questions dynamically using an AI model prompted with the company's own context, covering product knowledge, sales technique, objection handling, and process understanding. No two sessions are identical.
Managers can see where each new hire is in the training programme, which segments they have completed, and which assessments they have passed, without needing to ask candidates to self-report.
Engineering
Technical challenges
Keeping AI Questions Relevant
Generic AI question generation produces generic questions. We built the prompt layer around the company's actual product catalogue, pricing structure, common objections, and installation scenarios so every question feels like something a salesperson would genuinely encounter.
Enforcing Progression Without Frustrating Users
Locked segments can feel punishing if not designed carefully. We built the assessment and unlock logic to be firm enough to enforce the standard while giving users clear feedback on what they need to review before retrying.
Dual Experience in a Single Codebase
Building two meaningfully different experiences for two different user types within the same Flutter app required a clean role-based architecture that kept the codebase maintainable without the two experiences leaking into each other.
Automating the Scheduling Handoff
The trigger that fires when a new hire completes training needed to connect reliably to the company's existing scheduling workflow. Building this handoff cleanly so it worked every time without manual intervention was a key integration challenge.
Results
Business impact
Every person who joins goes through the same material in the same order and demonstrates comprehension before progressing. The quality of preparation no longer depends on who was available to train them. Managers are notified when someone is ready rather than asked to chase it, and the interview itself becomes more productive because the conversation can start at a higher level. For existing staff, the AI-generated questions remain relevant as long as the company context is kept updated, meaning the practice tool continues to deliver value months and years after the initial build.
Build With Us