For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
Y

Yaisel Dominguez Cordero

Senior Hardware & FPGA Engineer in Microelectronics, Aerospace, and R&D Systems

Spain flagCastilleja de la Cuesta, Spain
$25.00/hrIntermediateAppenClickworkerOneforma

Key Skills

Software

AppenAppen
ClickworkerClickworker
OneFormaOneForma
TolokaToloka
TelusTelus
Other

Top Subject Matter

Hardware Engineering & FPGA System Design
Neuromorphic Computing & Neural Networks
Power Electronics & Mixed-Signal Systems

Top Data Types

ImageImage
TextText
DocumentDocument

Top Task Types

Bounding Box
Classification
Text Generation
Red Teaming
Transcription
Evaluation Rating
Data Collection
Question Answering
Segmentation
RLHF

Freelancer Overview

AI training and data annotation specialist with hands-on experience across text, image, video, audio, and document-based workflows. Contributed to multiple AI projects involving bounding box annotation and structured data extraction from event posters, speaker diarization and speaker-based transcription, English-to-Spanish machine translation evaluation, large-scale Spanish text classification, parenting-response generation for conversational AI, transactional email data collection, and quality assessment of highly realistic video content based on lip movement, gesture naturalness, and visual artifacts. Also supported QA responsibilities in translation evaluation based on strong review performance. Additional experience includes advanced AI evaluation and red teaming in complex STEM domains such as mathematics, physics, and statistics. Worked on long-form tasks involving multi-dimensional response assessment, rubric creation, prompt refinement, and adversarial prompt design to expose reasoning failures in frontier models. Brings a strong combination of linguistic accuracy, analytical rigor, attention to detail, guideline adherence, and consistency across sensitive, high-complexity annotation and evaluation environments.

IntermediateEnglishSpanish

Labeling Experience

LIGHTHOUSE

TextClassification
Contributed to a large-scale text annotation project focused on native Spanish-language data. Responsibilities included annotating and labeling text according to detailed client-provided guidelines, with tasks such as identifying themes, sentiment, categories, and other linguistic features. The work required careful guideline interpretation, strong language sensitivity, consistent decision-making, and regular self-checks to maintain annotation accuracy and quality across large datasets.

Contributed to a large-scale text annotation project focused on native Spanish-language data. Responsibilities included annotating and labeling text according to detailed client-provided guidelines, with tasks such as identifying themes, sentiment, categories, and other linguistic features. The work required careful guideline interpretation, strong language sensitivity, consistent decision-making, and regular self-checks to maintain annotation accuracy and quality across large datasets.

2026 - Present

ANDROMEDA

TextData Collection
Contributed to a data collection project focused on email communications related to the purchase process of tangible consumer goods. Responsibilities included gathering and organizing purchase-related emails covering different stages of the transaction, such as order confirmation, payment updates, shipping notifications, delivery status, and other communications sent by both the seller and the carrier. All data handling followed privacy requirements, with personally identifiable information redacted through the application before submission. The work supported the creation of structured, privacy-compliant datasets for AI training and analysis.

Contributed to a data collection project focused on email communications related to the purchase process of tangible consumer goods. Responsibilities included gathering and organizing purchase-related emails covering different stages of the transaction, such as order confirmation, payment updates, shipping notifications, delivery status, and other communications sent by both the seller and the carrier. All data handling followed privacy requirements, with personally identifiable information redacted through the application before submission. The work supported the creation of structured, privacy-compliant datasets for AI training and analysis.

2026 - Present

SPIDEY

DocumentRed Teaming
Contributed to a highly complex AI red teaming project focused on mathematics, engineering, and physics tasks requiring advanced prompt design and file-based reasoning. Responsibilities included creating sophisticated prompts designed to expose model weaknesses by requiring the processing of complex input materials across three different file formats and the generation of at least two complex output files. The work also involved evaluating model reasoning with rubric-based criteria, with the objective of identifying cases where the model failed at least 60% of the reasoning evaluation. Tasks required careful prompt construction, rubric design and application, output validation, and the use of publicly available repository files with CC0 licensing. Each task was research-intensive and typically required around five hours to complete.

Contributed to a highly complex AI red teaming project focused on mathematics, engineering, and physics tasks requiring advanced prompt design and file-based reasoning. Responsibilities included creating sophisticated prompts designed to expose model weaknesses by requiring the processing of complex input materials across three different file formats and the generation of at least two complex output files. The work also involved evaluating model reasoning with rubric-based criteria, with the objective of identifying cases where the model failed at least 60% of the reasoning evaluation. Tasks required careful prompt construction, rubric design and application, output validation, and the use of publicly available repository files with CC0 licensing. Each task was research-intensive and typically required around five hours to complete.

2025 - Present

LYRA

AudioTranscription
Contributed to a speech annotation project focused on speaker identification, diarization consistency, and speaker-based transcription. Responsibilities included identifying individual speakers within audio recordings, maintaining speaker consistency throughout the full audio, and producing accurate transcriptions attributed to the correct speaker. Special attention was given to segmentation rules, including segment length constraints, precise start and end boundaries based on speaker turns, overlapping speech, and other temporal alignment requirements. All work was completed in accordance with detailed annotation guidelines to ensure high-quality training data for speech and language AI systems.

Contributed to a speech annotation project focused on speaker identification, diarization consistency, and speaker-based transcription. Responsibilities included identifying individual speakers within audio recordings, maintaining speaker consistency throughout the full audio, and producing accurate transcriptions attributed to the correct speaker. Special attention was given to segmentation rules, including segment length constraints, precise start and end boundaries based on speaker turns, overlapping speech, and other temporal alignment requirements. All work was completed in accordance with detailed annotation guidelines to ensure high-quality training data for speech and language AI systems.

2025 - Present

SONIC

VideoEvaluation Rating
Evaluated highly realistic video content for AI data annotation and quality assessment purposes. Tasks included reviewing and classifying videos based on multiple visual and behavioral parameters such as lip movement accuracy, gesture naturalness, visual consistency, the presence of glitches or rendering artifacts, etc. Worked with videos that could appear fully realistic regardless of whether they were AI-generated or not, applying project guidelines to assess overall quality and classification criteria consistently.

Evaluated highly realistic video content for AI data annotation and quality assessment purposes. Tasks included reviewing and classifying videos based on multiple visual and behavioral parameters such as lip movement accuracy, gesture naturalness, visual consistency, the presence of glitches or rendering artifacts, etc. Worked with videos that could appear fully realistic regardless of whether they were AI-generated or not, applying project guidelines to assess overall quality and classification criteria consistently.

2025 - Present

Education

U

University of Seville

Master of Science, Microelectronics

Master of Science
2011 - 2013
U

University of Havana (CUJAE)

Bachelor of Science, Telecommunication and Electronic Engineering

Bachelor of Science
2004 - 2009

Work History

S

Self-Employed

Business Owner

Castilleja de la Cuesta
2023 - Present
I

Integrasys

R&D Hardware Engineer

Camas
2023 - 2023