Jession Diwangan
Jession Diwangan
AI Engineer
GitHub LinkedIn Email Kaggle Huggingface

MFG Eng for Data Science
Jan 2023 - March 2024
Analog Devices, Inc Logo
As the lead developer, I led the creation of a Machine health monitoring system using a ensemble model trained trained on audio data to automate monitoring and prevent IC package cracks, deployed on Orangepi SBC.

Developed a Selenium-based automated allocation tool, reducing 98% of manual planning allocation and saving three days of work weekly, supporting planners across Philippines and Thailand. Deployed on enterprise servers.

Developed a vision system with yolov8 for automated IC counting in trays using Jetson Nano as the hardware, eliminating manual checks and identifying misoriented units.

Developed a RAG chatbot to enhance knowledge and information retrieval for equipment engineers, improving efficiency in equipment repairs and technical reviews.
Test Dev Eng
Nov 2019 - Jan 2023
Analog Devices, Inc Logo
Development, Debugging, and Evaluation of Test Solutions from Characterization to Release to Manufacturing for Simultaneous Sampling SAR ADC converters.

Built a Automated Flex Patterns Translator that transforms text pattern files, converting undocumented flex patterns into a clear, formatted output. Aimed at simplifying the pattern analysis process which significantly reduces the amount of debugging work from months to hours. It serves as a solution to manual pattern checking.

Trained a LGBM machine learning model to predict parameter shifts in components under temperature. The goal was to optimize manufacturing testing by eliminating unnecessary temperature stages, resulting in significant time savings.
Undergrad
2013 - 2019
xu logo
Took up BS Electronics and Communication Engineering. We built a wireless system for cashierless POS for our final year project. Graduated cum laude. Electronics Engineering Licensure Exam passer.


Pet Projects
Chat Icon
Developed noah, an advanced AI sales representative utilizing a multi-agent system. This innovative solution integrates Retrieval-Augmented Generation (RAG), web search capabilities, and a custom-built Python sandbox environment for complex calculations and code execution. Noah seamlessly combines these technologies to provide intelligent, data-driven sales support and problem-solving capabilities.

act logo
This web app allows you to train an audio classifier using ensemble models on any audio dataset. It is designed for non coders to build model automatically.
idwp logo
IDWYP is a python project takes an image and splits it into segments, each fitting on a specified paper size. The segments are then compiled into a single PDF file.
your-face logo
YOURFACE ai is a working desktop app that let you upload any facial images then get automatically recognized later.


Misc Unsorted
  • 32nd ASEMEP National Technical Symposium. Team ADI awarded as Best overall industry paper. See here.
  • SEAP 2018: My Experience. Short sharing.
  • Random shenanigans during college graduation. Article, Link.