AI-Powered Home Service Hiring Platform

A home service hiring platform that connects customers with service providers. It includes service requests, worker applications, document verification, bookings, ratings, and AI-assisted price recommendations based on relevant market factors.

View Code

homeservice.app

🏠

Home Service Platform

AI-assisted home service marketplace connecting customers with verified service providers.

Tech Stack

Languages

Kotlin

Mobile Development

Jetpack Compose Navigation Compose Material Design 3

Backend & API

REST API Authentication

Database Design

Room Database SQLite

AI / Machine Learning

Machine Learning Price Recommendation

Tools and Workflow

Git GitHub Android Studio Postman

🚀 Overview

The AI-Powered Home Service Hiring Platform is an end-to-end mobile marketplace that bridges the gap between customers seeking home services and skilled workers who can deliver them. Built with Kotlin and Jetpack Compose, the platform streamlines the entire service lifecycle—from posting a service request and reviewing worker applications to verifying documents, booking appointments, and leaving ratings after the job is complete.

A key differentiator of the platform is its AI-assisted price recommendation engine. By analyzing relevant market factors such as service type, location, worker experience, demand, and historical pricing, the model suggests fair and competitive prices that benefit both customers and service providers. Built with a modular architecture and REST API integration, the system is designed to scale as the marketplace grows.

✨ Key Features

Service Requests: Customers can post detailed service requests with location, budget, and preferred schedule.
Worker Applications: Service providers browse requests and apply for jobs that match their skills and availability.
Document Verification: Identity and skill-based documents are collected and verified to build trust on the platform.
Booking Management: Customers accept applications, schedule appointments, and track booking status in real time.
Ratings & Reviews: Both customers and workers can rate and review each other after a completed service.
AI Price Recommendations: Machine learning models suggest optimal prices based on market factors for informed decision-making.
RESTful API Architecture: A modular backend serves the Android app with clear separation of concerns and secure data flow.
Modern Android UI: Built with Jetpack Compose and Material Design 3 for a smooth, responsive, and accessible user experience.

💡 Technical Challenges

Challenge 1: Real-Time Booking & Status Synchronization

Problem: When a customer accepts a worker's application, the booking must immediately reflect across both the customer's and the worker's dashboards, while preventing double-booking of the same slot.

Solution: Implemented a state machine for booking lifecycle (pending → accepted → in-progress → completed) with a single source of truth on the backend. The Android app polls the REST API at a regular interval and refreshes the UI on status changes, with local caching via Room Database for offline resilience.

Challenge 2: AI Price Recommendation Accuracy

Problem: Predicting a fair service price depends on multiple correlated factors (location, worker experience, service type, demand), and early models produced inconsistent recommendations.

Solution: Trained a regression model on historical market data, encoding categorical features such as service type and location while normalizing numerical inputs. The model is served through the API, and the app displays the recommended price alongside the customer's budget so both parties can negotiate with data-backed context.

Challenge 3: Trust & Document Verification at Scale

Problem: A marketplace is only as good as its trust model—verifying worker identity and skills manually does not scale as the platform grows.

Solution: Designed a document upload and verification pipeline where workers submit ID and certification documents. Verified workers are marked with a trusted badge, and the ratings system reinforces accountability after every completed service, encouraging high-quality work.

📈 Outcome / Lessons Learned

Architecture & Engineering Excellence

Full-Stack Mobile Development: Built a complete marketplace spanning a Jetpack Compose frontend, REST API backend, and Room Database persistence.
ML Integration: Learned to integrate a machine learning model into a production mobile workflow, from data preprocessing to serving predictions via API.
Modular Design: Structured the app into independent features (requests, bookings, ratings, AI) that communicate through well-defined interfaces.

Product Impact

Smarter Pricing: AI recommendations help customers set realistic budgets and workers price their services competitively.
Trust Through Verification: Document verification and ratings reduce friction and build confidence between strangers on the platform.
End-to-End Workflow: Users can complete the entire service lifecycle—from request to rating—within a single polished app.

Real-World Learning

ML in Production: Discovered that model accuracy in training ≠ real-world performance—continuous feedback and retraining are essential.
Marketplace Dynamics: Balancing the needs of both customers and providers requires careful UX and pricing decisions.
Offline-First Thinking: Mobile users expect reliability, so caching with Room Database and handling network gracefully became a priority.