Matilda T.

Key Skills

Software

Other

Top Subject Matter

Chichewa/English audio transcription and annotation guidelines

Multilingual social media sentiment classification (X/Twitter)

Legal Services & Contract Review

Top Data Types

Audio

Text

Document

Top Task Types

Transcription

Emotion Recognition

Freelancer Overview

Audio Transcription & Annotation (Chichewa/English). Brings 7+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Other. Education includes Bachelor of Arts, University of Pretoria and Certificate in Artificial Intelligence Fundamentals, IBM SkillsBuild (2026). AI-training focus includes data types such as Audio and Text and labeling workflows including Transcription and Emotion Recognition.

Labeling Experience

Social Data Annotation & Sentiment Labeling

OtherTextEmotion Recognition

Performed social text data labeling for sentiment classification using structured analysis tooling. Validated and reviewed automatically generated sentiment labels for accuracy prior to publishing. Applied a quality-control mindset to maintain consistent labeled outputs for downstream analysis. • Assigned sentiment labels to X/Twitter text data using categorical tagging. • Reviewed model or tool outputs to correct inaccurate labels. • Used Python with VADER to support sentiment analysis and labeling. • Ensured labeling consistency through quality checks before final submission.

2026 - Present

Audio Transcription & Annotation (Chichewa/English)

OtherAudioTranscription

Applied detailed transcription guidelines to Chichewa and English audio samples, including verbatim vs. cleaned transcription. Labeled speaker turns and timestamp-aligned segments while marking uncertainty for ambiguous or unidentified speakers. Ensured consistent punctuation and bracket-tag formatting across disfluency removal and content preservation decisions using a structured rubric. • Transcribed and annotated non-verbal sounds (e.g., pauses, inaudible/unclear segments, foreign language). • Differentiated disfluencies to remove (false starts, stutters, filler repetition) versus content to preserve (profanity, grammatical errors). • Followed strict bracket-tagging conventions and language-specific style rules. • Applied edge-case judgment for ambiguous speakers with designated uncertainty markers.

2026

Education

Work History