Image Annotation
- Bounding Boxes
- Polygon Annotation
- Semantic Segmentation
- Instance Segmentation
- Object Detection
Senior AI Data Annotation & Evaluation Specialist
High-Quality Data Annotation & AI Evaluation
12+ years supporting large-scale AI training across Computer Vision, NLP/LLM, and multimodal systems — segmentation, keypoints, LLM evaluation, search relevance, and dataset QA.
12+Years experience
4Top AI vendors
QAProduction-level
Experience with leading AI platforms
AppenTELUS International AILionbridgeClickworker
About
About Alexandru
Alexandru-Catalin Ciobanu is a Senior AI Data Annotation & Evaluation Specialist with 12+ years supporting large-scale AI training pipelines across Computer Vision, NLP/LLM, and multimodal systems.
He has contributed to programs involving semantic segmentation, keypoints and headpose annotation, bounding boxes, video tracking, LLM response evaluation, hallucination and factuality checks, translation quality, search relevance, ads rating, and dataset QA — consistently meeting production-level accuracy and guideline-compliance standards.
Long-term contributor to programs with Appen, TELUS International AI, Lionbridge and Clickworker, working across multilingual datasets and multi-stage QC cycles.
EducationBSc IT, Univ. Craiova
Experience12+ years
Open toNew AI projects
Services
End-to-end annotation & evaluation
Production-grade data services for computer vision, LLMs, and multimodal AI.
Video Annotation
- Object Tracking
- Frame-by-Frame Annotation
- Action Recognition
- Motion Tracking
Computer Vision Datasets
- Dataset Preparation
- Data Labeling
- Quality Assurance
- Dataset Validation
AI & LLM Evaluation
- Prompt & Response Rating
- Hallucination & Factuality Checks
- Reasoning & Instruction Following
- AI Safety Evaluation
- Search & Ads Relevance
OCR & Data Validation
- OCR Review
- Transcription QA
- Linguistic Validation
- Multilingual Quality Control
Portfolio
Selected work & case studies
Representative project areas drawn from long-term contributions to leading AI data platforms. Client-specific details are kept confidential.
Computer Vision
⌗Semantic Segmentation
Polygon-based segmentation on multimodal CV programs, contributing to dataset quality used for model fine-tuning.
- Pixel-accurate polygon labels with strict guideline adherence
- Client-reviewed QC cycles to reach production-level quality
- Tooling: CVAT, SAM, MatrixGo
Computer Vision
◈Bounding Box Annotation
Object detection labeling across diverse imagery, with multi-stage QC and consistency control.
- High-volume bounding box and object detection labeling
- Consistency control across long-running batches
- Tooling: CVAT, Roboflow
Computer Vision
✦Keypoint & Pose Estimation
Human pose landmarks and headpose annotation supporting model training and evaluation.
- Body keypoints, facial landmarks, and headpose annotation
- Frame-by-frame consistency for video tracking
- Calibration against client gold sets
LLM / NLP
▱LLM Evaluation
Response rating across relevance, reasoning, factuality, hallucination, and safety dimensions.
- Prompt-response evaluation and side-by-side preference rating
- Hallucination and factual consistency checks
- Translation quality and instruction-following evaluation
Search & Ads
⌕Search Relevance & Ads Rating
Long-term search relevance and ads quality programs for production ranking systems.
- Query-result relevance grading at scale
- Ads quality and content evaluation
- Strict guideline compliance across multilingual datasets
QA
☑Dataset QA & Validation
Final-stage validation, consistency checks, and audio/text QA across multilingual training data.
- QA Final Validation on audio segmentation programs
- Linguistic QA, transcription review, and categorization
- Test-set validation supporting fine-tuning pipelines
Skills
Tools, techniques & domains
CV Tools
CVATRoboflowSAMMatrixGoAnnotation
Semantic SegmentationBounding BoxesKeypointsHeadposeVideo AnnotationAI Evaluation
LLM EvaluationSearch EvaluationAds RatingHallucination ChecksDomains
Computer VisionNLPMultimodal AIDataset QA Experience
12+ years powering AI teams
Long-term contributor on programs from the world's leading AI data platforms.
AI Data Annotation & Evaluation Specialist — Freelance
Appen · May 2023 – Present · RemoteMultimodal CV + NLP + Audio programs: segmentation, keypoints, headpose, translation quality, LLM response evaluation, and audio QA Final Validation.
AI Data Annotation & Evaluation Specialist — Freelance
Lionbridge → TELUS International AI · Mar 2013 – Present · RemoteLong-term LLM/NLP evaluation, search relevance, ads rating, content quality, and multilingual dataset improvement for production ranking and AI systems.
AI Data Annotation Contributor — Freelance
Clickworker · Mar 2023 – Present · RemoteDataset creation, categorization, transcription, linguistic QA, and consistency checks under tight turnaround and accuracy requirements.
Certifications
Continuous learning & credentials
EITC/AI/GCML – Google Cloud Machine Learning
EITCI Institute · 2026Supervised ML pipelines, training, evaluation, and scalable AI workflows on GCP.
EITC/CL/GCP – Google Cloud Platform
EITCI Institute · 2026GCP core services for AI: infrastructure, storage, APIs, IAM, and security.
EITC/CP/PPF – Python Programming Fundamentals
EITCI Institute · 2026Python 3 fundamentals and scripting for automation and AI tasks.
Artificial Intelligence Foundations: Machine Learning
LinkedIn Learning · 2026Core ML concepts and practical AI applications.
Hands-On Data Annotation: Applied Machine Learning
LinkedIn Learning · 2026Practical AI annotation workflows, dataset QA, and ML integration.
TELUS Digital – AI Community Certificate
TELUS International · 2024LLM evaluation, dataset quality, prompt rating, bias, reasoning, and hallucination checks.
AI Training – Level 1 Qualification
Freelancer Global FleetAI output evaluation, prompt alignment, model scoring. Score: 87%.
BSc in Information Technology
University of CraiovaAcademic foundation in information technology.
Why work with me
Built for serious AI teams
12+ Years of Industry Experience
Long-term contributor on programs from Lionbridge / TELUS, Appen, and Clickworker.
High Accuracy Standards
Pixel-precise labels and consistent ground truth your models can trust.
Fast Turnaround Times
Aggressive timelines without sacrificing the quality of your training data.
Scalable Annotation Workflows
Repeatable labeling, review, and QA processes for production datasets.
AI Training Data Understanding
Workflows aligned with the needs of ML, CV, NLP, and multimodal systems.
Production-Level Quality Assurance
Multi-stage QC, consistency checks, and guideline compliance.
Process
A proven five-step workflow
01
Project Discussion
Goals, data types, edge cases, and success criteria for your model.
02
Requirements Analysis
Annotation guidelines, taxonomies, and QA standards defined in detail.
03
Annotation & Evaluation
Production labeling and evaluation runs with calibrated review.
04
Quality Assurance
Multi-layer QA, consistency checks, and reviewer arbitration.
05
Final Delivery
Validated datasets and reports delivered in your preferred format.
Contact
Let’s Build Better AI
Tell me about your dataset, evaluation needs, or annotation pipeline. I’ll respond within one business day.
✉Emailoffice@altema.ro
✉Secondary emailalexandrucatalin.ciobanu04@gmail.com
☎Phone+40 761 615 598
inLinkedInalexandru-catalin-ciobanu
⌖LocationCraiova, Romania — Remote
Need high-quality training data for your AI model?
From Computer Vision annotation to LLM evaluation and dataset QA, Altema delivers reliable, production-ready AI data services.
Request a Quote Today