Anotace dat Jobs

We are looking for freelancers for audio annotation work in Welsh (United Kingdom). This is not simple transcription. Tasks include: Speaker segmentation Verbatim transcription Handling overlapping speech Sound tagging Following guidelines carefully Requirements Native or fluent in Welsh (United Kingdom) Experience in transcription or audio annotation Basic understanding of segmentation or tagging Good listening skills and attention to detail To apply please answer Agreed rate per minute Daily capacity Do you have transcription or annotation experience yes or no What languages and locales are you confident in Have you worked on segmentation overlap or tagging tasks yes or no Shortlisted candidates will receive guidelines and a test.

ANNOTATOR JOB DESCRIPTION — Dental X-ray AI Project Title: Dental X-ray Image Annotator (Remote, Freelance) Duration: 1–2 months (extension possible) Time Commitment: Approx. 2–3 hours/day Compensation: US$ 5 Per image About the Role: We are building an AI to detect dental pulp stones and root canal complexity from dental radiographs. You’ll be annotating X-ray images using a custom GUI tool (simple point-and-click, no technical skills required after training). Responsibilities: Annotate pulp stones and canal paths on provided dental X-rays Follow exact shape, boundary, and label protocol Submit JSON or project output for each image Maintain annotation quality and consistency Collaborate via WhatsApp for feedback Requirements: Back...

Description We are building a network of specialists who help validate and improve the reliability of enterprise AI systems. Our company, , works with organisations deploying AI assistants and knowledge systems powered by large language models and RAG (Retrieval-Augmented Generation). Your role will be to review AI responses, detect hallucinations, validate grounding against source material, and help create evaluation datasets. This is not model development. This role focuses on AI quality assurance and validation. Responsibilities • Review LLM responses for factual accuracy • Identify hallucinations and fabricated references • Verify whether answers are grounded in provided documents • Detect retrieval vs generation failures in RAG systems • Score responses...