Alex Mwangi

Key Skills

Software

Data Annotation Tech

Remotasks

Top Subject Matter

LLM Optimization and Safety Alignment

GenAI Model Coding & Reasoning Evaluation

LLM Prompt Engineering and Factual Verification

Top Data Types

Text

Image

Document

Top Task Types

Red Teaming

Prompt Response Writing SFT

Segmentation

Freelancer Overview

AI Operations Specialist (Contract). Brings 3+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Internal, Proprietary Tooling, and Data Annotation Tech. Education includes Bachelor of Science, Jomo Kenyatta University of Agriculture and Technology (2021). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Red Teaming, Computer Programming, and Coding.

Labeling Experience

AI Operations Specialist (Contract)

TextRed Teaming

As an AI Operations Specialist at Handshake AI, I led red-teaming efforts to uncover vulnerabilities in LLM safety filters. My work involved structured feedback and iterative evaluation for improved model robustness. I also helped design internal labeling rubrics to enhance annotation consistency across teams. • Led comprehensive red-teaming workflows targeting LLM safety features. • Developed structured, actionable feedback loops for iterative improvement. • Architected guidelines that reduced inter-annotator disagreement by 15%. • Curated synthetic datasets for technical domain gap areas.

As an AI Operations Specialist at Handshake AI, I led red-teaming efforts to uncover vulnerabilities in LLM safety filters. My work involved structured feedback and iterative evaluation for improved model robustness. I also helped design internal labeling rubrics to enhance annotation consistency across teams. • Led comprehensive red-teaming workflows targeting LLM safety features. • Developed structured, actionable feedback loops for iterative improvement. • Architected guidelines that reduced inter-annotator disagreement by 15%. • Curated synthetic datasets for technical domain gap areas.

2024 - Present

LLM Specialist (GenAI & Coding)

As an LLM Specialist at Outlier, I debugged and optimized complex code snippets generated by advanced language models. I also refined chain-of-thought reasoning outputs, ranking responses to support reward model training. My role was crucial to ensuring accuracy and helpfulness in coding and reasoning tasks for AI model improvement. • Reviewed and corrected model-generated Python and JavaScript code. • Evaluated 'Chain of Thought' responses for logical integrity. • Performed comparative analysis for reward modeling (HHH). • Enhanced multi-step logical reasoning outputs for LLMs.

As an LLM Specialist at Outlier, I debugged and optimized complex code snippets generated by advanced language models. I also refined chain-of-thought reasoning outputs, ranking responses to support reward model training. My role was crucial to ensuring accuracy and helpfulness in coding and reasoning tasks for AI model improvement. • Reviewed and corrected model-generated Python and JavaScript code. • Evaluated 'Chain of Thought' responses for logical integrity. • Performed comparative analysis for reward modeling (HHH). • Enhanced multi-step logical reasoning outputs for LLMs.

2023 - 2024

Core Data Contributor

Data Annotation TechTextPrompt Response Writing SFT

At DataAnnotation.tech, I engineered prompts to test LLM creativity and factual recall capabilities. I fact-checked outputs and conducted quality audits on peer-generated data to bolster dataset fidelity. These efforts aimed to minimize hallucinations and maintain data integrity for RAG systems. • Designed complex prompt scenarios for SFT-based evaluation. • Verified model outputs against trusted primary sources. • Executed peer audits on high-priority annotation projects. • Enhanced accuracy and reliability of AI-generated datasets.

At DataAnnotation.tech, I engineered prompts to test LLM creativity and factual recall capabilities. I fact-checked outputs and conducted quality audits on peer-generated data to bolster dataset fidelity. These efforts aimed to minimize hallucinations and maintain data integrity for RAG systems. • Designed complex prompt scenarios for SFT-based evaluation. • Verified model outputs against trusted primary sources. • Executed peer audits on high-priority annotation projects. • Enhanced accuracy and reliability of AI-generated datasets.

2023 - 2023

AI Model Trainer

RemotasksImageSegmentation

At Remotask, I executed pixel-precise semantic segmentation for computer vision training datasets. My work involved detailed object annotation and analysis of systemic model errors to inform targeted fine-tuning. The goal was to maximize dataset accuracy for robust machine learning. • Annotated objects with pixel-level precision in computer vision tasks. • Identified recurring errors and presented findings to engineers. • Supported dataset optimization for AI fine-tuning protocols. • Contributed data-driven insights for improved CV model performance.

At Remotask, I executed pixel-precise semantic segmentation for computer vision training datasets. My work involved detailed object annotation and analysis of systemic model errors to inform targeted fine-tuning. The goal was to maximize dataset accuracy for robust machine learning. • Annotated objects with pixel-level precision in computer vision tasks. • Identified recurring errors and presented findings to engineers. • Supported dataset optimization for AI fine-tuning protocols. • Contributed data-driven insights for improved CV model performance.

2021 - 2022

Education

Work History