N Akash Rao

Key Skills

Software

AWS SageMaker

Anno-Mage

Appen

Argilla

Axiom AI

Clickworker

CloudFactory

Data Annotation Tech

Datatroniq

Datasaur

Datature

Deep Systems

Doccano

Google Cloud Vertex AI

HiveMind

Humanatic

Img Lab

Labelbox

Lionbridge

OneForma

OpenCV AI Kit (OAK)

Snorkel AI

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

Bounding Box

Classification

Computer Programming Coding

Evaluation Rating

RLHF

Freelancer Overview

I am an AI enthusiast with hands-on experience in data labeling, LLM evaluation, and computer vision annotation. My background includes projects in LLM-powered threat modeling, prompt-based AI evaluation, and multimodal data training. I have worked extensively with tools like CVAT, Labelbox, Prodigy, and Doccano for structured labeling of text, image, and code datasets. I also bring strong experience in prompt engineering, rubric-based scoring, classification, bounding box annotation, and real-time data analysis. My recent work includes developing a ResNet-based visual defect detection system during Google Summer of Code and designing an AI-driven threat prediction model using OpenAI and Mistral. I’ve labeled and evaluated text responses for LLM safety and performance, applied autoencoders for anomaly detection, and built STEM tutoring tools using retrieval-augmented generation (RAG). With a deep focus on precision, clarity, and domain understanding, I’m confident in contributing to high-quality AI training across vision, NLP, and security-driven applications.

Labeling Experience

Financial Sentiment Analysis & Fraud Detection Annotation

OneformaTextSegmentationClassification

Worked on labeling financial news articles and transaction data for AI models focused on sentiment analysis and fraud detection. This included classifying articles as positive, neutral, or negative and tagging transactions as either legitimate or suspicious. The work required attention to context, patterns in language, and risk indicators. Helped train models to automatically detect market sentiment shifts and flag potential fraud. This project highlighted the importance of contextual understanding in financial text, and our annotations contributed directly to improved model precision and faster threat detection.

Worked on labeling financial news articles and transaction data for AI models focused on sentiment analysis and fraud detection. This included classifying articles as positive, neutral, or negative and tagging transactions as either legitimate or suspicious. The work required attention to context, patterns in language, and risk indicators. Helped train models to automatically detect market sentiment shifts and flag potential fraud. This project highlighted the importance of contextual understanding in financial text, and our annotations contributed directly to improved model precision and faster threat detection.

2024 - 2024

Medical Image Annotation for AI Diagnostics

LabelboxImageBounding BoxSegmentation

Contributed to the development of an AI-powered diagnostic tool by annotating medical images such as X-rays, MRIs, and CT scans. Used bounding boxes and segmentation techniques to accurately label regions of interest (e.g. tumors, lesions, abnormalities). Worked closely with annotation leads and quality teams to ensure clinical-grade accuracy. Our efforts led to a 25% boost in model performance, improving diagnostic reliability and speed. This project demanded high attention to detail, consistency, and familiarity with medical imaging types.

Contributed to the development of an AI-powered diagnostic tool by annotating medical images such as X-rays, MRIs, and CT scans. Used bounding boxes and segmentation techniques to accurately label regions of interest (e.g. tumors, lesions, abnormalities). Worked closely with annotation leads and quality teams to ensure clinical-grade accuracy. Our efforts led to a 25% boost in model performance, improving diagnostic reliability and speed. This project demanded high attention to detail, consistency, and familiarity with medical imaging types.

2023 - 2023

Education

Work History