Sandra Elaine B.

Key Skills

Software

Don't disclose

Scale AI

Appen

Top Subject Matter

NLP annotation for conversational AI (NER, intent, sentiment, coreference, dialogue acts)

Autonomous vehicle perception and computer vision annotation (text plus image/video labeling workflows)

NLP annotation for search/virtual assistant and supporting audio transcription/voice labeling

Top Data Types

Text

3D Sensor

Image

Document

Top Task Types

Entity (NER) Classification

Segmentation

Question Answering

Classification

Freelancer Overview

Senior AI Data Annotation Consultant | Upwork (Freelance), Remote. Brings 23+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, Scale AI, and Appen. Education includes Doctor of Philosophy, Northwestern University (2004) and Master of Science, University of Illinois at Urbana-Champaign (1999). AI-training focus includes data types such as Text and 3D Sensor and labeling workflows including Entity (NER) Classification, Segmentation, and Question Answering.

Labeling Experience

Senior AI Data Annotation Consultant | Upwork (Freelance), Remote

Don't discloseTextEntity (NER) Classification

Provided end-to-end NLP data labeling and annotation services for enterprise clients, including conversational AI datasets. Labeled and classified text data for tasks such as named entity recognition, intent classification, sentiment analysis, coreference resolution, and dialogue act labeling. Developed labeling taxonomies and ran multi-pass quality audits using inter-annotator agreement metrics to ensure consistent, high-precision outputs.• Labeled 500,000+ text data points across multiple NLP classification and extraction tasks.• Maintained ~98.7% labeling accuracy through structured guideline and QA processes.• Reduced inter-annotator disagreement by up to 35% via detailed annotation guidelines and taxonomies.• Performed multi-pass quality audits using Cohen's Kappa, maintaining IAA scores above 0.85.

2020 - Present

Data Annotation Team Lead | Scale AI, Remote (Chicago, IL)

Scale AI3D SensorSegmentation

Led a remote data annotation team delivering high-accuracy labels for autonomous vehicle perception training data. Oversaw computer-vision labeling of image frames using bounding boxes and segmentation approaches, supporting model training for robotics and automotive applications. Built annotator training and introduced an automated-plus-human tiered quality review pipeline to reduce label error rates over time.• Achieved 99.2% label accuracy on LiDAR point cloud-related datasets for perception systems.• Annotated 2M+ image frames using bounding boxes, semantic segmentation masks, and instance segmentation.• Reduced onboarding time by 40% and improved new-hire productivity by 28% within 30 days via training programs.• Reduced label error rates from 4.1% to 0.8% over six months using a tiered quality review pipeline with automated rule checks and human review.

2018 - 2019

NLP Data Specialist | Appen, Remote (Chicago, IL)

AppenTextQuestion Answering

Completed high-volume NLP data annotation for search engine and virtual assistant applications. Labeled and reviewed text for sentiment analysis, topic classification, textual entailment, and question-answer pair generation. Also supported audio-related labeling efforts and contributed feedback to improve annotation rubrics and label consistency across client projects.• Annotated and reviewed 300,000+ English text samples with consistent quality and speed.• Labeled sentiment, topics, textual entailment, and QA pairs for client NLP training datasets.• Labeled speech data for accent identification, speaker diarization, and emotion detection with 97%+ transcription accuracy.• Assisted evaluation and ranking of AI-generated responses to support early-stage RLHF pipeline development.

2016 - 2018

Research Data Analyst (Graduate Research) | University of Chicago, Dept. of Cognitive Science

Don't discloseTextClassification

Designed and managed a longitudinal linguistic corpus consisting of 50,000+ annotated text samples for research on semantic ambiguity resolution. Applied statistical analysis tools to evaluate inter-rater reliability and validate annotation consistency across a team of research assistants. Developed annotation frameworks and contributed to scholarly dissemination through peer-reviewed publications and conference presentations.• Built and maintained a 50,000+ annotated corpus for doctoral research on semantic ambiguity resolution.• Evaluated annotation consistency and inter-rater reliability using R, SPSS, and Python.• Co-authored annotation frameworks used in subsequent AI and NLP research (cited 140+ times).• Presented findings at ACL and Cognitive Science Society conferences to support community adoption of the annotation methods.

2004 - 2012

Education

Work History