Post

Project Showcase and Technical Summaries

Real projects in ML, backend, and automation with measurable outcomes.

Project Showcase and Technical Summaries

This page is organized for quick screening with a consistent format:

  • Duration
  • Tech
  • Impact
  • Links

Project Cards

WBC Analyzer: AI-Powered Pathology Assistant

  • Duration: Jan 2026 - Apr 2026
  • Tech: DenseNet121, Flask REST API, Grad-CAM, GPT-4o, Gemini
  • Impact:
    • Achieved 98.5% validation accuracy and 89.6% OOD performance on unseen hardware data.
    • Delivered approximately 200 ms average API response time for near real-time usage.
    • Improved explainability with Grad-CAM and multi-model LLM review flows.
  • Links:
    • GitHub: Available on request
    • Demo: Available on request

Popcorn Wagon: Hybrid Movie Recommendation System

  • Duration: Jan 2025 - May 2025
  • Tech: TruncatedSVD, TMDB API, MovieLens, Pandas, Dask, Spotify Annoy
  • Impact:
    • Built a hybrid recommendation strategy combining collaborative and content-based methods.
    • Processed large datasets with a scalable ETL pipeline.
    • Reduced recommendation latency to millisecond level using approximate nearest neighbors.
  • Links:
    • GitHub: Available on request
    • Demo: Available on request

Marketplace Automation Core

  • Duration: Dec 2025 - Mar 2026
  • Tech: Python, Appium, Selenium, UiAutomator2
  • Impact:
    • Built a modular automation core with config-driven flow definitions.
    • Separated reusable execution logic from platform-specific selectors.
    • Improved stability with retry handling, scroll adjustments, and reusable execution patterns.
  • Links:
    • GitHub: Available on request
    • Demo: Private

Portal Cleaner Ultimate: Desktop RPA Suite

  • Duration: Jun 2025 - Jul 2025
  • Tech: Python, Selenium, Tkinter, Multithreading, HTML/JS test harness
  • Impact:
    • Reduced daily manual workload by 90%+ (4-6 hours to under 30 minutes).
    • Built a local mock ERP environment to run offline end-to-end reliability tests.
    • Implemented dynamic retry and centralized error logging for fault-tolerant execution.
    • Maintained responsive UI under heavy background ingestion and scraping tasks.
  • Links:
    • GitHub: Available on request
    • Demo: Internal

CV Download

  • PDF CV: Please contact me if you want the latest signed PDF version.

Professional Experience Highlights

Operations Manager and Python Developer - zidorun (Jun 2025 - Present)

  • Founded and operated a niche collectibles micro-venture across multiple C2C marketplaces.
  • Engineered Python backend workflows and sync pipelines for inventory and pricing.
  • Processed 700+ transactions with 99% success rate.
  • Reduced repetitive manual operations by 90%+.

Intern Engineer - Platech Coating (Jun 2025 - Jul 2025)

  • Developed Python-based desktop automation suite using Selenium and Tkinter.
  • Engineered multithreaded GUI architecture and fault-tolerant retry workflows for unstable ERP portal interactions.
  • Built report generation, order frequency analysis, and HTML-based product image extraction scripts.

Detailed case-study posts with architecture diagrams and API snippets will be published next.

This post is licensed under CC BY 4.0 by the author.