Alex M.

Key Skills

Software

Appen

Labelbox

Top Subject Matter

Large Language Model alignment

safety evaluation

and preference-data generation

Top Data Types

Text

Audio

Image

Top Task Types

RLHF

Function Calling

Classification

Freelancer Overview

Senior AI Trainer & RLHF Specialist (Scale AI). Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Scale AI Nucleus, Appen Connect, and Labelbox. Education includes Master of Science, Carnegie Mellon University (CMU) (2021) and Bachelor of Science, University of Nairobi (2018). AI-training focus includes data types such as Text, Audio, and Geospatial and labeling workflows including RLHF, Function Calling, and Evaluation.

Labeling Experience

Senior AI Trainer & RLHF Specialist (Scale AI)

TextRLHF

Built and evaluated high-quality preference datasets for RLHF pipelines to support frontier LLM fine-tuning. Conducted hallucination audits and factuality assessments to identify errors and produce corrected exemplars for supervised fine-tuning datasets. Led adversarial red teaming to surface safety vulnerabilities, jailbreak patterns, and model failure modes. • Preference data creation for reasoning, coding, and instruction-following tasks • Hallucination/factuality assessment with corrected exemplars for SFT • Adversarial red teaming and safety evaluation across production LLMs • Reasoning step verification and multi-step output review for math and code generation

2023 - Present

AI Data Trainer – NLP & Multilingual (Appen)

TextFunction Calling

Performed large-scale NLP annotation and quality review across text classification, NER, sentiment analysis, and intent detection. Led multilingual Swahili and Kikuyu training projects for speech recognition and text generation models targeting East African markets. Executed function-calling and agent-simulation workflows to test LLM-powered agents for customer service, scheduling, and retrieval use cases. • Text classification, NER, sentiment analysis, and intent detection • Multilingual dataset creation for Swahili and Kikuyu • Function calling and agent simulation evaluation for practical workflows • Contractor training and guideline updates to reduce labeling errors

2021 - 2023

AI Trainer & Data Annotator (Sama)

Don't discloseAudio

Completed multimodal annotation for autonomous-vehicle and medical-imaging AI projects. Produced speech and audio labeling outputs including transcriptions, speaker emotion tags, and ASR validation for English and Swahili datasets. Participated in LLM evaluation by rating model responses on helpfulness, harmlessness, and honesty criteria. • Image annotation: bounding boxes, semantic segmentation, and keypoints • Speech/audio labeling: transcription, emotion tagging, ASR validation • LLM response evaluation (helpfulness/harmlessness/honesty) • Time series labeling for financial forecasting and anomaly detection

2020 - 2021

Junior AI Data Labeler (Techbridge Africa / Ajira Digital)

Classification

Performed foundational AI data annotation tasks across text categorization, image tagging, and video frame labeling for local and international AI clients. Assisted with point cloud annotation for 3D object detection models used in smart city infrastructure projects across East Africa. Supported satellite imagery labeling for geospatial AI applications such as land use, vegetation, and infrastructure feature identification. • Text categorization and image tagging • Video frame labeling • Point cloud labeling for 3D object detection • Satellite imagery labeling for land use/vegetation/infrastructure

2018 - 2019

Education

Work History