Skip to content
NLNuvyntra Labs
All R&D projects

R&D project

Vehicle Inspection (Kotlin)

Offline-first inspection with AI damage marking and branded PDF/JSON reports

Kotlin + Jetpack Compose Android app: configurable checklist, VIN OCR, Gemini/manual damage marking, and branded PDF + JSON reports. Offline-first; Firebase and AI are optional.

Kotlin

Abstract

Vehicle Inspection is an offline-first Android app, built with Kotlin, Jetpack Compose, and Clean Architecture. It guides inspectors through a configurable checklist, captures per-item photos with Gemini Vision and manual damage marking, and produces a branded report (JSON + PDF).

Problem

Certified resale without a structured, versioned inspection is marketing copy. Paper checklists and chat photos cannot gate inventory or survive questionnaire edits offline.

Solution

A checklist-first capture product: identify the vehicle, inspect against a vendor questionnaire, grade each item, mark damage, then export an in-app report plus offline-rendered PDF and structured JSON.

Architecture

  • Kotlin, Jetpack Compose, Clean Architecture.
  • Room + encrypted file storage as the source of truth; WorkManager syncs to Firebase when online.
  • Questionnaire snapshot pinned on each inspection.
  • PDF rendered offline from an HTML/CSS template via WebView.

Capabilities

  • Configurable checklist filtered by New vs Old vehicle applicability.
  • VIN entry or on-device OCR, decode, and classification.
  • Gemini Vision damage boxes plus manual annotations.
  • Branded multi-section PDF and JSON export.

Audience

Inspectors and dealer ops who need a decision-ready report — including AssureCars publish gates.

Outcomes

  • Inspection is a versioned product, not a sidecar spreadsheet.
  • Works fully offline; cloud and AI are additive, not required.

Related r&d projects