Rifki Rosada | Public Android + AI

Scanberry

Created a mobile computer vision prototype for fruit scanning workflows with edge inference constraints.

Mobile Computer Vision Engineer 4-week prototype Public product build
  • Problem

    The workflow needed rapid classification support in environments with unstable connectivity.

  • My scope

    Built a lightweight model iteration and evaluation loop.

  • Result

    Validated a practical on-device scan flow for classification without cloud dependency.

Problem

The workflow needed rapid classification support in environments with unstable connectivity.

Constraints

  • Inference latency needed to remain acceptable on-device
  • Model behavior had to tolerate imperfect capture angles
  • UI had to keep scan interactions simple for repeat use

Delivery

  • Built a lightweight model iteration and evaluation loop.
  • Implemented mobile capture and inference triggers.
  • Tested the scan-and-review flow for repeated use in constrained conditions.

Result

  • Delivered a demo-ready scan workflow with local inference.
  • Identified clear next steps for dataset and model improvement.
  • Documented architecture tradeoffs for deployment planning.

Tech stack

  • Computer Vision
  • Android
  • TensorFlow Lite
  • Mobile UX

Related delivery proof across internal tools, automation, Android + AI, and supporting product systems.

Public Android + AI

Public build

OffScan AI

Android Product Engineer6-week build and release

Published an Android workflow that performs OCR locally without cloud upload requirements.

  • Problem

    Users needed OCR extraction with stronger privacy guarantees and no network dependency.

  • My scope

    Designed an offline-first OCR flow for camera capture and image import.

  • Result

    Published an Android workflow that performs OCR locally without cloud upload requirements.

  • Android
  • On-device OCR
  • Edge AI
Read case study

Public Android + AI

Public build

I-ScanTea

ML Prototype Engineer5-week prototype

Demonstrated an on-device classification workflow for field capture scenarios.

  • Problem

    Manual quality checks required repeated visual comparison and lacked a fast digital workflow.

  • My scope

    Prepared the dataset and labeling workflow for the target classes.

  • Result

    Demonstrated an on-device classification workflow for field capture scenarios.

  • TensorFlow Lite
  • Android
  • Python
Read case study

Android + AI

Client delivery

Media Android App - AI Chat + Search UX

Android AI UX Engineer3-week delivery sprint

Made the AI-assisted discovery flow clearer and more reliable from chat prompt to search result.

  • Problem

    Users lacked a clear transition from AI chat prompts to meaningful search actions, and state sync issues reduced UX reliability.

  • My scope

    Worked inside the existing WebView-based AI entry path so the feature fit the live app architecture.

  • Result

    Made the AI-assisted discovery flow clearer and more reliable from chat prompt to search result.

  • Android
  • Kotlin
  • WebView
Read case study

Planning a similar build?

Share the workflow, delivery risk, and timeline. I will reply with the best starting scope.

Share your scope Email project brief