Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am an analytical AI Training Specialist with a focus on red teaming, adversarial prompting, and high-precision data annotation. I have proven expertise in stress-testing large language models for ethical alignment, safety, and logical consistency, and I bring experience in specialized domain consulting for beauty and retail sectors. I excel in RLHF, prompt engineering, and rigorous data validation, including transcription and multimodal evaluation. I thrive in contract and freelance environments, delivering high-quality outputs and actionable feedback to improve model behavior and safety.…I am an analytical AI Training Specialist with a focus on red teaming, adversarial prompting, and high-precision data annotation. I have proven expertise in stress-testing large language models for ethical alignment, safety, and logical consistency, and I bring experience in specialized domain consulting for beauty and retail sectors. I excel in RLHF, prompt engineering, and rigorous data validation, including transcription and multimodal evaluation. I thrive in contract and freelance environments, delivering high-quality outputs and actionable feedback to improve model behavior and safety.

Ashanti Harris

AI Engineer, Data Scientist, Data Analyst, +2





I am an analytical AI Training Specialist with a focus on red teaming, adversarial prompting, and high-precision data annotation. I have proven expertise in stress-testing large language models for ethical alignment, safety, and logical consistency, and I bring experience in specialized domain consulting for beauty and retail sectors. I excel in RLHF, prompt engineering, and rigorous data validation, including transcription and multimodal evaluation. I thrive in contract and freelance environments, delivering high-quality outputs and actionable feedback to improve model behavior and safety.…I am an analytical AI Training Specialist with a focus on red teaming, adversarial prompting, and high-precision data annotation. I have proven expertise in stress-testing large language models for ethical alignment, safety, and logical consistency, and I bring experience in specialized domain consulting for beauty and retail sectors. I excel in RLHF, prompt engineering, and rigorous data validation, including transcription and multimodal evaluation. I thrive in contract and freelance environments, delivering high-quality outputs and actionable feedback to improve model behavior and safety.

Available to hire

I am an analytical AI Training Specialist with a focus on red teaming, adversarial prompting, and high-precision data annotation. I have proven expertise in stress-testing large language models for ethical alignment, safety, and logical consistency, and I bring experience in specialized domain consulting for beauty and retail sectors.

I excel in RLHF, prompt engineering, and rigorous data validation, including transcription and multimodal evaluation. I thrive in contract and freelance environments, delivering high-quality outputs and actionable feedback to improve model behavior and safety.

Skills

Prompt Engineering

AI Data Labelling

AI Collection (Video)

Experience Level

Prompt Engineering

Expert

AI Data Labelling

Expert

AI Collection (Video)

Intermediate

Language

English

Fluent

Work Experience

AI Safety & Alignment Specialist at Remotasks

January 1, 2023 - July 1, 2023

Executed Adversarial Training on the Pi Platform, utilizing advanced prompting techniques to identify model vulnerabilities and break the bot. Audited model responses against safety and policy rubrics to ensure enterprise-level ethical standards. Provided high-signal feedback on tone, style, and factual accuracy, contributing to refinement of LLM conversational logic. Conducted rigorous peer reviews of training data to maintain 98%+ accuracy across project batches.

AI Training Specialist / Data Specialist (Contract) at Outlier (Scale AI)

November 1, 2025 - Present

Reinforcement Learning from Human Feedback (RLHF): Ranked and refined AI-generated responses to improve model accuracy and conversational flow. Prompt Engineering: Developed complex prompts to evaluate LLM performance across parameters (logic, creativity, technical accuracy). Fact-Checking & Truthfulness: Verified AI outputs against primary sources to mitigate hallucinations and ensure factual integrity.

Professional Transcriptionist at Rev

April 1, 2021 - May 1, 2022

Transcribed high-complexity audio and video sessions for general conversations. Maintained a consistent on-time delivery rate of 99% while managing high-volume, deadline-driven task queues.