About

Ahmad Arif Sultoni

AI Engineer

I am an AI Engineer with 2 years of experience in developing deep learning and computer vision-based solutions. I have a great interest in optimizing machine learning models for various tasks such as object detection, segmentation, and image classification, especially in the mining industry sector. With a strong background in Python, TensorFlow, and PyTorch, I always strive to deliver the best results in every project while remaining focused on efficient use of resources. I enjoy working in a team and am open to continuing to learn and develop in the field of AI, with the aim of making a positive contribution in every available opportunity.

From: Pemalang, Central Java, Indonesia
Lives In: Jakarta, Indonesia
Age: 25
Gender: Male

Skills

Python
C++
TensorFlow
OpenCV
PyTorch
Linux
Google Colab
Google Cloud
SQL
NVIDIA Jetson
Natural Language Processing
Computer Vision
OCR
Image Processing
Embedded Systems

Education

Mechatronic Engineering Education 2022

Universitas Negeri Yogyakarta - Bachelor's degree

GPA: 3.52

Essay:

Development of Object Detection Prototype Using TensorFlow Object Detection API and Google Colaboratory as Learning Media in Visual Sensing Robot Course

1st place Indonesian wheeled soccer Robot Contest at the National Level (2021)
3rd place Indonesian wheeled soccer Robot Contest at the National Level (2020)
2nd place Indonesian wheeled soccer Robot Contest at the Regional Level (2019)

Experience

Artificial Intelligence Engineer, PT. Safepedia Global Teknologi (04/2022 - Present)
- AI Core Logging - PT. Freeport Indonesia
  - Developed a deep learning-based application for analyzing core photos from drilling operations.
  - Detected and analyzed key features on core images with segmentation and joint analysis.
  - Differentiated between natural and mechanical fractures based on images.
  - Built a user-friendly GUI to simplify input and user interaction.
  - Reduced RAM usage in processing deep learning models to improve performance.
- GeoAI - Mind ID
  - Developed a mining land detection system using satellite imagery data and a custom YOLOv8-seg based segmentation model.
  - Measured mining area sizes with precise geospatial analysis.
  - Leveraged Google Colab, Makesense.AI, and segment-geospatial for data annotation, segmentation, and processing.
- Geologging - PT. Antam
  - Built a quicklog application powered by deep learning for analyzing core photos from gold mining exploration.
  - Applied Detectron2 for core detection and segmentation, accelerating the geological logging process.
  - Developed deep learning application methods to increase effectiveness with limited resources.
- Grain Counting Analysis - PT. Timah
  - Prepared a mineral rock image dataset to train a segmentation model with Detectron2.
  - Detected, counted rock types, and analyzed mineral content using predefined constants.
  - Optimized the model for resource efficiency and high accuracy in mineral analysis.

Freelance Mentor, Indobot Academy (10/2021 - 01/2022)
- Assisted students of all skill levels in core subjects, including programming and electronics.
- Delivered online course material about IoT.
- Trained adult learners in classes, workshops, and via remote learning programs.

Computer Vision Programmer, UNY Robotics Team (10/2018 - 10/2021)
- Developed an object detection system on a wheeled soccer robot using Python programming language, OpenCV library, and TensorFlow framework on Jetson Nano.