Siddhartha Sinha

Key Skills

Software

Appen

Clickworker

Data Annotation Tech

Mercor

OneForma

Remotasks

Telus

Other

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Computer Programming/Coding

Data Collection

Evaluation/Rating

Mapping

Object Detection

Freelancer Overview

AI Trainer and Full-Stack Engineer with 4 years of hands-on experience evaluating, training, and improving Large Language Models using prompt engineering, data annotation, and dataset QA. Skilled in analysing LLM outputs for accuracy, safety, reasoning quality, and factual correctness. Strong technical background in Python, NLP frameworks, and cloud tools, enabling high-quality dataset creation and model behaviour refinement.

Labeling Experience

ui-images , gauge-annotation, correct-counting, aspect-ratio-cropping, ig-entity-tagging, hard-grounding-samples, group-person-removal , asr-transcript-accuracy-v2, website-generation, add-perception-labels

Scale AIImagePoint Key PointClassification

Annotated and labeled images for UI elements, gauges, objects, people, and webpages, including tasks like counting, cropping, entity tagging, perception labeling, and accuracy verification to support AI model training and visual understanding.

Annotated and labeled images for UI elements, gauges, objects, people, and webpages, including tasks like counting, cropping, entity tagging, perception labeling, and accuracy verification to support AI model training and visual understanding.

2021

stationary-camera-transforms, engaging-videos, ig-entity-tagging-videos

TelusVideoPoint Key PointClassification

Worked on video annotation tasks involving tagging video content, identifying key objects/actions, and providing quality ratings based on clarity, relevance, and engagement. Ensured accurate labeling and consistent evaluation to support model training and improve video understanding performance.

Worked on video annotation tasks involving tagging video content, identifying key objects/actions, and providing quality ratings based on clarity, relevance, and engagement. Ensured accurate labeling and consistent evaluation to support model training and improve video understanding performance.

2021

Education

Work History