Mildred S.

Key Skills

Software

Appen

Clickworker

iMerit

Img Lab

Label Studio

Mercor

Micro1

Lionbridge

Mindrift

OneForma

Remotasks

SuperAnnotate

Toloka

Telus

Top Subject Matter

Languages

linguistics

e-commerce

legal

conversational AI. I have also worked across autonomous vehicle datasets

Top Data Types

Image

Video

Text

Top Task Types

Bounding Box

Entity (NER) Classification

Segmentation

Relationship

Question Answering

Classification

Freelancer Overview

Over the past five years, I've worked remotely across some of the largest AI data platforms in the industry; TELUS International, RWS, CrowdGen, and OneForma. Handling the full range of data labelling and AI training tasks. On the annotation side, that has included bounding box and polygon labelling for computer vision datasets, NER and intent classification for NLP pipelines, sentiment and toxicity scoring, and search relevance rating for major search engine clients. I've worked across image, video, text, and audio modalities, and consistently maintained accuracy scores above 98% on QA audits with inter-annotator agreement tracked at kappa 0.87 on larger batch projects. On the AI training side, I've contributed to RLHF workflows for large language models. Writing and ranking prompt-response pairs, running safety and red-team evaluations, and flagging policy violations with documented rationale. My linguistics background (BA and MA) has been particularly useful on tasks that go beyond surface labelling, including semantic role labelling, coreference resolution, and evaluating machine-translated output for fluency and adequacy. I'm comfortable picking up new annotation guidelines quickly and have been assigned to higher-complexity task pools across multiple platforms as a result.

Labeling Experience

Education

Work History