Justin A.

Key Skills

Software

Other

Mercor

Micro1

Labelbox

Internal/Proprietary Tooling

Don't disclose

Top Subject Matter

Generative AI

Multimodal AI Output

Search Relevance and AI Safety

Top Data Types

Audio

Video

Text

Document

Top Task Types

RLHF

Classification

Prompt + Response Writing (SFT)

Segmentation

Bounding Box

Object Detection

Text Generation

Question Answering

Text Summarization

Data Collection

Evaluation/Rating

Transcription

Freelancer Overview

Gen AI Analyst. Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other, Mercor, and Micro1. Education includes Master of Science, Strayer University and Bachelor of Arts, University of Memphis. AI-training focus includes data types such as Audio, Video, and Text and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

Gen AI Analyst

OtherAudioRLHF

As a Gen AI Analyst at Black Crown Nexus, I evaluated and refined multimodal datasets including audio, visual, and text to improve next-generation AI models. I conducted Reinforcement Learning from Human Feedback (RLHF) to minimize hallucinations and bolster context recognition. Model outputs across various formats were rigorously analyzed to ensure fidelity and adherence to safety guidelines. • Refined and validated AI outputs for accuracy and contextual relevancy. • Identified and addressed edge cases in audio-visual data pipelines. • Collaborated with cross-disciplinary teams to implement best practices in data annotation. • Enhanced machine learning models through quality assurance of diverse data types.

2023 - Present

Video Diarization Expert

Micro1VideoSegmentation

At Micro1 as a Video Diarization Expert, I enhanced multi-speaker video content quality through advanced labeling and audio segmentation strategies. The role included segmenting multi-speaker audio, visual speaker labeling, and managing challenging diarization cases like overlapping speech. My work enabled users to filter and navigate video archives more efficiently. • Implemented innovative labeling for multi-speaker videos. • Reduced cognitive load via audio and visual segmentation. • Spearheaded Conversation Intelligence platform integration. • Handled edge cases in diarization for seamless user navigation.

2026 - 2026

Tool Use Expert

MercorTextClassification

During my term as a Tool Use Expert at Mercor, I performed diagnostic labeling across a seven-tool ecosystem. The primary objective was to ensure accuracy in intent classification, reducing errors and hallucinations within tool recommendations. This role directly supported the reliability and performance of AI-driven tool selection systems. • Conducted intensive diagnostic labeling sessions. • Prioritized accuracy to minimize wrong-tool usage. • Collaborated on intent classification scheme refinement. • Documented labeling processes for consistency.

2026 - 2026

Generalist Writing Expert

MercorTextPrompt + Response Writing (SFT)

As a Generalist Writing Expert at Mercor, I engineered complex prompt sets to stress-test Large Language Model (LLM) capabilities. The focus was on aligning model outputs with human intent and maintaining high UX quality. My work contributed to prompt-response datasets used in training advanced LLM systems. • Developed and evaluated a wide variety of LLM prompts. • Assessed generative outputs for UX compliance. • Documented prompt-performance trends for analysis. • Ensured intent alignment through prompt engineering best practices.

2026 - 2026

AI Safety Analyst

OtherTextClassification

At Global Logic, I served as an AI Safety Analyst transforming complex search relevancy data into actionable quality strategies. My duties included auditing content for Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T) and conducting consistency audits to refine query-response performance. I helped ensure alignment with quality guidelines and safe search ecosystems. • Audited and rated content for high-stakes domain trustworthiness. • Identified and mitigated misleading or harmful material. • Documented technical solutions for query-response gaps. • Enhanced overall content relevance and safety.

2026 - 2026

Show more

Education

Work History