Soundhariya Muthukumar

Motivated and detail-oriented Data Annotation Specialist with 7+ years of hands-on experience in image, text, audio, and video annotation, AI training data preparation, prompt engineering, and team leadership. Seeking to leverage deep expertise in data labeling workflows, JSON structuring, and quality assurance to contribute meaningfully to cutting-edge AI and machine learning projects.

AI Training Data Contributor at Label Your Data (2026-02 – Present)

Contributed to AI model training by generating structured natural language prompts and well-formed JSON datasets for bathroom layout planning, enabling spatial reasoning capabilities in LLMs.

• Designed prompts across diverse bathroom types: half bath, compact, guest, master, accessible, narrow, and luxury.
• Crafted spatial instructions covering fixture types, room dimensions, wall placement, and fixture orientation.
• Developed varied complexity levels — from minimal 2-fixture layouts to 10+ fixture luxury configurations.
• Ensured prompts captured real-world design constraints including accessibility clearance and logical door/window placement.
• Created structured JSON training files pairing prompts with bathroom layout data: boundary coordinates, fixture placement, and architectural elements.
• Defined and populated fixture metadata fields — type, coords, and orientation (UP/DOWN/LEFT/RIGHT).
• Organized room boundary coordinates as polygon arrays for room sizes ranging from 140×200 cm to 460×320 cm.
• Maintained 100% schema consistency and validated JSON structure across 10+ diverse bathroom layout entries.

AI Data Annotation Intern at Klatch Technologies (2025-08 – 2025-09)

Annotated and classified complex multimodal datasets to support model training across Computer Vision and NLP projects.

• Executed object detection and semantic segmentation assignments with meticulous attention to detail and consistent quality.
• Collaborated with the annotation team, contributing to project milestones through strong work ethic and accuracy.

Data Annotation Specialist & Audio Transcriber at Freelancer – Upwork, TELUS (2022-02 – Present)

• Annotated images, text, video, and audio using industry-standard tools including LabelBox and CVAT across diverse domains.
• Performed audio transcription and text annotations for complex legal documents with high accuracy requirements.
• Transcribed audio sermons with a focus on accuracy, speaker clarity, and contextual understanding (2+ years experience).
• Delivered high-quality bounding box annotations for varied image datasets including solar panels, house roofs, vehicles, and signage.

Team Lead – Data Annotation at Globistic Info Tech (2017-11 – 2020-02)

• Led a team of annotators, ensuring accuracy, consistency, and timely delivery across all annotation projects.
• Provided onboarding training and conducted quality control checks for new team members.
• Completed 6-month intensive hands-on training in OCR (Optical Character Recognition), Proofreading, and ePub conversion (2019).
• Supervised XML EPUB tagging projects, maintaining consistency and quality across large-scale digital content pipelines.
• Managed multiple data entry projects simultaneously, balancing speed, accuracy, and client requirements.

B.E. in Computer Science & Engineering – Christian College of Engineering & Technology (2016)

HSE – St. Joseph's Girls' Higher Secondary School (2012)

SSLC – St. Joseph's Girls' Higher Secondary School (2010)