Ayalkibet A.

Key Skills

Software

CVAT

Label Studio

Doccano

Top Subject Matter

Multilingual African language data (English, Amharic, Tigrinya, Oromifa) for LLM training and evaluation

Machine learning training pipeline and reproducible dataset workflows

Risk-level image classification for emergency alert systems

Top Data Types

Text

Image

Top Task Types

Data Collection

Object Detection

Classification

Tracking

RLHF

Red Teaming

Transcription

Freelancer Overview

Founder — AI Data Labeling & Annotation Startup (AfriData Solutions). Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include CVAT, Label Studio, and Doccano. Education includes Bachelor of Science, Adama Science and Technology University (2025). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Translation, Localization, and Data Collection.

Labeling Experience

Founder — AI Data Labeling & Annotation Startup — AfriData Solutions, Addis Ababa

CVATImageClassificationTracking

No description provided.

2025 - Present

Founder - AfriData Solutions

CVATImageClassificationTracking

Founded and led an AI data labeling and annotation startup focused on building African-context machine learning training datasets. Owned end-to-end workflow setup, client outreach, and multilingual dataset quality processes across text, image, and speech modalities. Leveraged Python/SQL-style data thinking, tooling like CVAT, Label Studio, and Doccano, and strong data validation practices to deliver usable evaluation-ready data. • Built annotation workflows for multilingual and multimodal data using CVAT, Label Studio, and Doccano • Designed data quality and inter-annotator consistency checks for Amharic, Tigrinya, Oromifa, and English • Supported business development via outreach to AI platforms and research institutions • Coordinated operational roadmap and technical stack implementation for scalable labeling and evaluation work

2025 - Present

Founder — AI Data Labeling & Annotation Startup (AfriData Solutions)

CVATText

Led an African-context AI training-data business focused on multilingual multilingual dataset preparation and localization across English, Amharic, Tigrinya, and Oromifa. Established end-to-end annotation workflows using CVAT, Label Studio, and Doccano with structured review and quality validation. Coordinated data quality processes including validation, review, and inter-annotator consistency checks.• Set up multilingual annotation pipelines and documentation for client delivery.• Implemented validation and review steps to ensure label consistency.• Supported outreach and client acquisition for labeling and evaluation engagements.• Managed dataset design decisions for training data readiness.

2025 - Present

Data & AI Analyst (Project) - Decentralized Emergency Alert System

CVATImageClassification

Worked on a final-year project analyzing large image datasets for risk-level classification using computer vision models. Evaluated model performance statistically to validate accuracy and optimize results for deployment readiness. Built backend services with a web API to integrate and serve analysis outputs and documented the methodology for the project team. • Classified image records into Low, Medium, and High risk levels using YOLOv8 and CLIP-based validation • Conducted statistical performance evaluation to measure accuracy and improve model outcomes • Developed FastAPI backend services to provide real-time analyzed data • Documented methodology to support team understanding and reproducibility

2024 - 2025

Data & AI Analyst — Decentralized Emergency Alert System (Final Year Project)

ImageObject Detection

Performed image dataset analysis by classifying records into risk levels using a trained YOLOv8 model with CLIP-based validation. Conducted statistical performance evaluation to validate accuracy and optimize model results. Built FastAPI backend services to integrate and serve analyzed data for real-time use by the team.• Applied model-based validation using CLIP to support classification reliability.• Measured and evaluated performance using statistical methods and KPI tracking.• Implemented real-time data-serving APIs with FastAPI.• Documented the methodology for team reproducibility.

2024 - 2025

Show more

Education

Work History