Javier J.

Key Skills

Software

OneForma

Label Studio

Supervisely

Data Annotation Tech

CVAT

Top Subject Matter

LLM evaluation

hallucination detection

RLHF quality assurance

Top Data Types

Text

Video

Audio

Document

Image

Top Task Types

RLHF

Transcription

Data Collection

Text Generation

Freelancer Overview

Tier 3 AI Evaluation Specialist (RLHF Expert). Brings 7+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Internal and Proprietary Tooling. Education includes Bachelor of Science, Universidad Autónoma de Tlaxcala (2020) and Mechatronics Technician, CECyTE 11 (2023). AI-training focus includes data types such as Text, Video, and Audio and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

AI Integration Projects (Conoce tu municipio)

DocumentText Generation

Developed an AI integration project that produced digital, QR-based informational content. Created AI-generated avatars and accompanying materials to improve public access to historical information. Organized content outputs to support a public-facing educational experience. • Built “Conoce tu municipio” AI-generated avatar and QR content materials • Generated and packaged text/visual content for historical information access • Structured outputs for public delivery and comprehension • Supported deployment of AI-driven educational materials

Present

Cultural Data Enrichment

TextData Collection

Curated culturally focused datasets intended to reduce geographic bias and improve local relevance. Selected and enriched examples reflecting regional culture and social environments. Used the resulting data to better align AI outputs with cultural context. • Curated regional culture and social-environment datasets • Reduced geographic bias through targeted data enrichment • Improved local relevance of training/evaluation data • Supported cultural alignment for downstream AI use

Present

Audio Transcription & Refinement

AudioTranscription

Reviewed and corrected AI-generated transcriptions to improve ASR training datasets. Evaluated transcription quality and refined text outputs for better accuracy and usability. Contributed to data preparation for automatic speech recognition model improvement. • Assessed AI transcriptions for errors and needed corrections • Refined transcription text to strengthen ASR training inputs • Improved consistency and correctness of speech-to-text data • Supported creation of higher-quality training sets

Present

Multimodal Evaluation (Tier 3)

Video

Performed multimodal comparative evaluation by ranking outputs across video, image, and audio. Assessed alignment with technical constraints and user intent using structured evaluation methods. Supported iterative improvements by identifying mismatches between generation behavior and expected criteria. • Compared and ranked multimodal outputs from Video, Image, and Audio systems • Checked alignment with technical constraints and intended user outcomes • Used evaluation criteria to surface weaknesses for downstream improvements • Supported model assessment workflows for multimodal quality

Present

Tier 3 AI Evaluation Specialist (RLHF Expert)

TextRLHF

Led Tier 3 RLHF evaluation work focused on identifying and correcting model failures. Audited factuality and hallucinations in LLM outputs to ensure grounded, accurate responses. Validated complex prompts against technical, ethical, and safety requirements. • Conducted hallucination and factuality audits on generated LLM text • Performed final validation (QA) for high-complexity prompts • Designed multidimensional Likert-scale rubrics to rate feasibility, truthfulness, and utility • Ensured outputs complied with strict safety and ethics guidelines

Present

Education

Work History