Athresh V K

Key Skills

Software

Mindrift

Appen

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Text

Computer Code Programming

Image

Audio

Top Label Types

Classification

Question Answering

Evaluation Rating

Computer Programming Coding

Object Detection

Prompt Response Writing SFT

Audio Recording

Transcription

Function Calling

Freelancer Overview

I am a detail oriented Business Analyst and AI data annotation specialist with hands on experience across multiple AI training and evaluation projects, including Mindrift (Apricon response evaluation), Alignerr, TAU, Nexus PL/SQL data curation, VLM SFT annotation, and ATC transcription. My work focuses on assessing AI model outputs for instruction following, truthfulness, harmlessness, completeness, and conversational quality, as well as creating structured prompts, validating SQL/PL SQL logic, and ensuring strict adherence to annotation guidelines and QA frameworks. Alongside my AI training work, I bring several years of professional experience in requirements analysis, quality assurance, SQL based validation, Agile delivery, and user acceptance testing across enterprise environments. This background strengthens my ability to deliver highly accurate, guideline compliant annotations, perform root cause analysis, and produce structured, reliable training data that improves AI model performance and consistency.

Labeling Experience

Conversational AI Evaluation – TAU / Multi-turn QA

MindriftTextEvaluation RatingFunction Calling

Reviewed multi turn AI conversations for coherence, tone handling, instruction compliance, and user intent understanding. Delivered structured feedback and preference judgments to enhance conversational quality and safety.

Reviewed multi turn AI conversations for coherence, tone handling, instruction compliance, and user intent understanding. Delivered structured feedback and preference judgments to enhance conversational quality and safety.

2025

SQL & PL/SQL Data Curation

AppenComputer Code ProgrammingEvaluation RatingComputer Programming Coding

Curated and validated SQL/PL-SQL prompts, queries, triggers, and stored procedures for AI training datasets. Ensured logical correctness, adherence to schema constraints, and compliance with multi level QA review guidelines.

Curated and validated SQL/PL-SQL prompts, queries, triggers, and stored procedures for AI training datasets. Ensured logical correctness, adherence to schema constraints, and compliance with multi level QA review guidelines.

2025

AI Response Evaluation

MindriftTextClassificationQuestion Answering

Evaluated AI generated responses across multiple quality dimensions including instruction following, truthfulness, harmlessness, completeness, conciseness, and dialogue naturalness. Applied strict annotation rubrics and provided structured rationales to improve model alignment and reliability.

Evaluated AI generated responses across multiple quality dimensions including instruction following, truthfulness, harmlessness, completeness, conciseness, and dialogue naturalness. Applied strict annotation rubrics and provided structured rationales to improve model alignment and reliability.

2025

ATC Transcription & Cleanup – Alignerr

LabelboxAudioClassificationAudio Recording

Performed high-accuracy transcription and normalization of Air Traffic Control communications. Applied aviation phraseology standards and QA review to ensure clarity, correctness, and dataset usability for speech recognition model training.

Performed high-accuracy transcription and normalization of Air Traffic Control communications. Applied aviation phraseology standards and QA review to ensure clarity, correctness, and dataset usability for speech recognition model training.

2025 - 2025

Multimodal Annotation – VLM SFT

MindriftImageObject DetectionEvaluation Rating

Annotated multimodal tasks requiring interpretation of screenshots, UI elements, and structured reasoning. Ensured guideline compliance, reasoning accuracy, and correct complexity classification across multiple coding and analysis scenarios.

Annotated multimodal tasks requiring interpretation of screenshots, UI elements, and structured reasoning. Ensured guideline compliance, reasoning accuracy, and correct complexity classification across multiple coding and analysis scenarios.

2025 - 2025

Education

Work History