Open to opportunities

MICHAEL CRISS

@michaelcriss

Message

AI trainer and software developer focused on secure systems, LLM evaluation, and dataset quality.

United States

Message

What I'm looking for

I seek roles where I can design rigorous evaluation pipelines, curate high-quality multimodal datasets, and improve LLM safety within collaborative, security-focused teams.

I am an AI trainer and software developer with experience building secure backend services and designing rigorous evaluation frameworks for large language models. I have developed secure APIs, internal tools for alignment testing, and structured test scenarios to measure instruction-following accuracy and identify edge-case behaviors.

My annotation and QA work includes producing 1,000+ multimodal annotations and 500+ high-quality audio recordings, achieving zero rejected data batches across multiple projects and reducing QA corrections by 35%. I collaborate closely with AI engineers and QA teams to deliver standardized datasets (CSV, JSON, WAV) compatible with model pipelines.

With academic research and teaching experience in computational linguistics, I have published peer-reviewed work on discourse clarity and trained others in annotation best practices. I seek roles where I can apply systems design, evaluation methodology, and dataset curation to improve model reliability and safety.

Experience

Work history, roles, and key accomplishments

Current

AI Research Assistant

Current

Outlier AI

Jan 2023 - Present (2 years 8 months)

Produced 500+ high-quality audio recordings for multimodal AI datasets, reduced QA corrections by 35% through improved annotation adherence, and developed style guides and workflows to streamline dataset delivery in CSV/JSON/WAV formats.

Audio Recording QA Processes CSV JSON WAV Formatting Speech To Text Pipelines

Current

AI Trainer

Current

Unknown

Jan 2023 - Present (2 years 8 months)

Reviewed and rated LLM outputs for tone, fluency, and factual accuracy, crafted training prompts and model responses, and produced high-quality annotated datasets while flagging hallucinations and bias.

Prompt Engineering Annotating Quality Assurance Content Moderation

Software Developer

Unknown

Jan 2022 - Dec 2024 (2 years 11 months)

Developed and maintained secure APIs and backend services in Python integrating with core banking systems and Temenos T24, delivered microservices in Agile sprints, and built internal tools to test LLM alignment and measure instruction-following accuracy.

Python Linux Administration API Development Microservices Agile Scrum Security Compliance Bash Scripting

Graduate Research Associate

Arizona State University

Conducted research on semantics and speech clarity for human–AI systems, built annotated voice–image datasets, published two peer-reviewed articles, and informed NLP preprocessing and data quality metrics.

Annotation Experiment Design Data Quality Multimodal Datasets

Graduate Teaching Fellow

Arizona State University

Taught undergraduate courses in semantics, phonology, and discourse analysis, designed assignments emphasizing voice clarity and annotation precision, and supervised undergraduate thesis projects.

Teaching Curriculum Design Student supervision Communication

Education

Degrees, certifications, and relevant coursework

Arizona State University

Linguistics and Communication

2017 - 2021

Activities and societies: Graduate Research Associate; Graduate Teaching Fellow; supervised undergraduate thesis projects; delivered workshops on speech clarity and annotation.

Completed studies in linguistics and communication with coursework and research focused on voice recording, transcription, and annotation tools supporting multimodal AI datasets.