Madhan K.

Key Skills

Software

Don't disclose

CVAT

Top Subject Matter

LLM response evaluation and multimodal dataset QA (text, image, audio)

Computer vision annotation (2D/3D) and dataset QA

Video annotation and dataset quality assurance

Top Data Types

Text

Image

Top Task Types

Segmentation

Tracking

Transcription

Cuboid

Bounding Box

Point/Key Point

Polyline

Freelancer Overview

Analyst–AI/LLM Practice at Innodata Inc.. Brings 4+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose and CVAT. Education includes Bachelor of Engineering, KSR College of Engineering (2023). AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Segmentation.

Labeling Experience

AI/LLM Analyst - Innodata

CVATImage

Evaluated and optimized LLM-generated responses to improve consistency, quality, and instruction-following performance. Reviewed and validated AI outputs across multimodal datasets spanning text, images, and audio while following quality guidelines. Used prompt analysis, response validation, and error analysis to support reliable datasets for model training and benchmarking. • LLM evaluation and response optimization • Multimodal AI output validation (text, image, audio) • Prompt analysis and instruction-following refinement • Dataset quality assurance for training and benchmarking

2025 - 2026

Analyst–AI/LLM Practice at Innodata Inc.

Don't discloseText

Evaluated and optimized LLM-generated responses to improve response consistency and overall output quality across text-based prompts and answers. Validated AI outputs against quality guidelines and performed error analysis to refine evaluation workflows. Produced high-quality validated datasets that support AI training, benchmarking, and evaluation pipelines.•Assessed instruction-following accuracy and consistency of model responses.•Reviewed and validated multimodal AI outputs for text components.•Performed response validation and error analysis to identify quality gaps.•Delivered datasets aligned to QA and training requirements.

2025 - 2026

Senior AI/Computer Vision Analyst - Objectways Technology

CVATImageBounding BoxSegmentation

Worked on large-scale multimodal AI dataset development spanning image, video, audio, and LiDAR (3D point cloud) data. Led annotation-oriented QA workflows, including structured review processes to reduce inconsistencies and improve dataset reliability. Maintained high annotation accuracy through validation checks and attention to detail across 2D and 3D labeling tasks. • 2D and 3D dataset QA workflows • Segmentation and bounding box labeling support • Polyline and keypoint labeling validation • Quality improvement through detailed review processes

2023 - 2025

Senior Analyst – AI/Computer Vision & Data Annotation at Objectways Technology

CVATImageSegmentationTracking

Performed large-scale 2D and 3D data annotation for multimodal AI datasets using structured QA workflows to maintain high labeling accuracy. Labeled visual content with segmentation and bounding-style annotations while reducing inconsistencies through detailed review processes. Ensured dataset readiness for downstream model training and computer vision workflows.•Created segmentation labels for image data.•Annotated bounding boxes and keypoints as part of computer vision labeling tasks.•Applied structured validation steps to reach and maintain 95%+ accuracy.•Conducted reviews and quality checks to reduce annotation inconsistencies.

2023 - 2025

Education

Work History