Precious Chepchirchir

Key Skills

Software

Telus

Other

Appen

Top Subject Matter

AI Data Evaluator & Multilingual Quality Analyst (Swahili–English, Legal Domain)

Multilingual Audio and Image Dataset Annotation

Swahili-English Audio Transcription

Top Data Types

Text

Audio

Image

Document

Top Task Types

Classification

Transcription

RLHF

Evaluation/Rating

Data Collection

Freelancer Overview

I have specialized in optimizing AI models through high-fidelity data annotation, linguistic evaluation, and transcription. My work consistently focuses on refining machine learning datasets by identifying subtle cultural nuances, grammatical inconsistencies, and tone shifts that are critical for natural language processing. What distinguishes my profile is the legal precision I bring to every task. As a current Law student (LLB), I apply rigorous analytical standards to data auditing, ensuring that every label or transcription meets strict project guidelines and ethical requirements.

Labeling Experience

AI Language Evaluator (Swahili-English)

TelusText

Validated Swahili-English sentence pairs for integration into AI training datasets, ensuring linguistic accuracy and cultural relevance. Assessed translation quality and reviewed the contextual tone for AI systems. Provided structured feedback to improve machine learning outputs and overall dataset usability. • Evaluated over 1,000 bilingual sentence pairs for model optimization. • Analyzed and flagged inconsistencies in translation outputs. • Enhanced dataset quality through rigorous language assessment processes. • Contributed to cross-cultural AI development projects.

Validated Swahili-English sentence pairs for integration into AI training datasets, ensuring linguistic accuracy and cultural relevance. Assessed translation quality and reviewed the contextual tone for AI systems. Provided structured feedback to improve machine learning outputs and overall dataset usability. • Evaluated over 1,000 bilingual sentence pairs for model optimization. • Analyzed and flagged inconsistencies in translation outputs. • Enhanced dataset quality through rigorous language assessment processes. • Contributed to cross-cultural AI development projects.

2025 - 2026

Annotation Contractor

AudioClassification

Categorized and annotated multilingual audio and image samples for machine learning training datasets. Classified audio and visual data based on predefined project criteria to uphold quality standards. Supported speech recognition system development through diverse dataset enhancement. • Labeled 300+ samples across multiple data modalities. • Applied strict quality control checks to all annotations. • Improved machine learning inputs for audio recognition tasks. • Consistently met project productivity benchmarks under remote conditions.

Categorized and annotated multilingual audio and image samples for machine learning training datasets. Classified audio and visual data based on predefined project criteria to uphold quality standards. Supported speech recognition system development through diverse dataset enhancement. • Labeled 300+ samples across multiple data modalities. • Applied strict quality control checks to all annotations. • Improved machine learning inputs for audio recognition tasks. • Consistently met project productivity benchmarks under remote conditions.

2025 - 2025

Freelance Transcriber

OtherAudioTranscription

Transcribed Swahili and English audio recordings, ensuring high levels of accuracy and consistency. Managed precise proofreading and time coding to meet stringent transcription standards. Addressed a wide range of dialects and complex speech patterns for training language models. • Processed over 100 hours of multilingual audio data. • Maintained consistency across both languages and formats. • Enhanced transcription fidelity for AI and speech systems. • Ensured reliable time-coding for downstream machine learning models.

Transcribed Swahili and English audio recordings, ensuring high levels of accuracy and consistency. Managed precise proofreading and time coding to meet stringent transcription standards. Addressed a wide range of dialects and complex speech patterns for training language models. • Processed over 100 hours of multilingual audio data. • Maintained consistency across both languages and formats. • Enhanced transcription fidelity for AI and speech systems. • Ensured reliable time-coding for downstream machine learning models.

2023 - 2024

Education

Work History