SYS_ARCH_PORTFOLIO
Deployment Status: Stable

NAOL
BERHANU

Applied AI & Data Systems Engineer. Building real-world systems with data pipelines, machine learning, and deployable applications.

Origin_Protocol

Addis Ababa-based applied AI engineer from Ethiopia, focused on turning data systems, OCR, and machine learning into usable tools with real-world impact.

Section_01 // Identity

Technical Summary

Applied AI engineer focused on building real-world systems combining data pipelines, machine learning, and deployable applications. Experience in OCR for low-resource languages, computer vision, and multi-sector data platforms. Contributing to open-source AI initiatives for Ethiopian languages and data accessibility.

Applied AI & Data Systems

Developing robust end-to-end ML pipelines and distributed data processing systems.

Low-Resource NLP

Advancing OCR and language modeling for Ethiopic scripts including Amharic and Afan Oromo.

Open Source AI

Contributing to accessible research and dataset democratization in the AI space.

System Architecture

Designing decoupled, microservice-oriented infrastructures for AI deployments.

Section_02 // Tech_Stack

Python
PyTorch
OpenCV
Pandas
NumPy
Flutter
SQL
Docker
Git
Laravel
AWS
Apache Spark

Section_03 // Deployments

CNN_MODEL_VERSION_2.0_KERNELS_3x3_STRIDE_1_PADDING_SAME_RELU_MAXPOOL_FLATTEN_DENSE_SOFTMAX_PLANT_DISEASE_DETECTION_METRIC_ACCURACY_0.985

Plant Disease Detection

ML // CV

Scalable ML system utilizing CNNs for early crop disease detection. Optimized for low-latency inference on edge devices.

#Python#PyTorch#OpenCV
DATASET_ET_AFAN_OROMO_AMHARIC_SCRAPING_CLEANING_TOKENIZATION_TRANSFORMER_READY_CORPUS_SIZE_500GB_OPEN_SOURCE_RESEARCH_ETHIOPIAN_AI

Dataset.et

DATA // NLP

Research platform for Ethiopic datasets. Focused on Amharic and Afan Oromo corpus collection and benchmark development.

#NLP#Pandas#Scrapy
OCR_ENGINE_TESSERACT_CUSTOM_TRAINING_LSTMS_IMAGE_PREPROCESSING_SKEW_CORRECTION_AMHARIC_SCRIPT_RECOGNITION_95_PERCENT_ACCURACY

Amharic OCR

CV // TOOLS

OCR pipeline for Amharic script. Integrates Tesseract engine with custom OpenCV image preprocessing layers.

#C++#Tesseract#SQL

Future Projects

ENQUEUED_IN_BACKLOG

Section_04 // Credentials

Addis Ababa University

Bachelor of Science in Computer Science

Jun 2021 - Jul 2024

Udacity

Artificial Intelligence & Data Analysis Nanodegree Programs

Oct 2025

Cisco Networking Academy

CCNAv7: Networking Fundamentals and Enterprise Networking

Aug 2024

Section_05 // Signal

Initiate Connection

Currently open to collaborative research, engineering consulting, and technical opportunities in the Applied AI space.