Vikash Bairwa

Key Skills

Software

Scale AI

Internal/Proprietary Tooling

CVAT

Other

Top Subject Matter

Computer Vision

LLM Training

Model Evaluation

Top Data Types

Text

Audio

Geospatial Tiled Imagery

Image

Top Label Types

Translation Localization

RLHF

Fine Tuning

Evaluation Rating

Prompt Response Writing SFT

Audio Recording

Question Answering

Tracking

Routing

Mapping

Classification

Transcription

Freelancer Overview

Senior AI Data Annotator & Prompt Engineer. Core strengths include CVAT and Other. AI-training focus includes data types such as Image and Audio and labeling workflows including Classification and Transcription.

Labeling Experience

Senior Hindi Data Annotator & Prompt Engineer

Scale AITextTranslation LocalizationRLHF

Multilingual Annotation & LLM Evaluation Platform: Outlier.AI | Tools: Remotasks, Outlier Software Duration: March 2024 – July 2024 Worked on a multilingual annotation project to improve chatbot accuracy, fluency, and cultural relevance for Hindi and multilingual users. The scope included prompt creation, AI response evaluation, quality rating, and response rewriting to enhance LLM performance. Tasks: Created localized prompts, evaluated and ranked responses for instruction adherence, fluency, coherence, factual accuracy, and tone, provided structured ratings with error identification, and rewrote outputs for clarity and native-level quality. Project Size: Processed a high volume of multilingual prompts and responses across diverse task categories. Quality Measures: Followed strict evaluation criteria, including factual verification, hallucination detection, safety compliance, and cultural appropriateness checks.

Multilingual Annotation & LLM Evaluation Platform: Outlier.AI | Tools: Remotasks, Outlier Software Duration: March 2024 – July 2024 Worked on a multilingual annotation project to improve chatbot accuracy, fluency, and cultural relevance for Hindi and multilingual users. The scope included prompt creation, AI response evaluation, quality rating, and response rewriting to enhance LLM performance. Tasks: Created localized prompts, evaluated and ranked responses for instruction adherence, fluency, coherence, factual accuracy, and tone, provided structured ratings with error identification, and rewrote outputs for clarity and native-level quality. Project Size: Processed a high volume of multilingual prompts and responses across diverse task categories. Quality Measures: Followed strict evaluation criteria, including factual verification, hallucination detection, safety compliance, and cultural appropriateness checks.

2022 - 2025

AI Hindi and Hinglish PodCast Recording

Internal Proprietary ToolingAudioTranslation LocalizationFine Tuning

AI Hindi and Hinglish Audio Training Platform: Babel Audio Role: AI Hindi and Hinglish Audio Trainer Duration: Aug 2025 – Jan 2026 Worked on conversational AI training by creating Hindi and Hinglish speech datasets to improve ASR and voice interaction quality. Tasks: Recorded role-based dialogues, ensured clear pronunciation and natural speech flow, reviewed audio for clarity and noise, and supported transcription alignment for AI training. Project Size: Produced 70+ conversational audio recordings across diverse real-life scenarios. Quality Measures: Followed guidelines for audio clarity, linguistic accuracy, and cultural authenticity to ensure high-quality training data.

AI Hindi and Hinglish Audio Training Platform: Babel Audio Role: AI Hindi and Hinglish Audio Trainer Duration: Aug 2025 – Jan 2026 Worked on conversational AI training by creating Hindi and Hinglish speech datasets to improve ASR and voice interaction quality. Tasks: Recorded role-based dialogues, ensured clear pronunciation and natural speech flow, reviewed audio for clarity and noise, and supported transcription alignment for AI training. Project Size: Produced 70+ conversational audio recordings across diverse real-life scenarios. Quality Measures: Followed guidelines for audio clarity, linguistic accuracy, and cultural authenticity to ensure high-quality training data.

2025 - 2025

Search Quality Rater (Hindi)

Internal Proprietary ToolingGeospatial Tiled ImageryQuestion AnsweringTracking

Search Quality Rater (Hindi) Platform: WeLocalize Duration: Feb 2024 – Nov. 2025 Evaluated search results and ads to improve relevance and user experience for Hindi-speaking users. Tasks: Rated results for intent match, relevance, linguistic accuracy, and webpage quality; assessed ads and landing pages; flagged misleading or low-quality content; provided structured feedback to improve rankings. Project Size: Processed a high volume of search queries across diverse topics and user intents. Quality Measures: Followed strict rating guidelines, ensured cultural and language accuracy, maintained consistency, and adhered to confidentiality standards.

Search Quality Rater (Hindi) Platform: WeLocalize Duration: Feb 2024 – Nov. 2025 Evaluated search results and ads to improve relevance and user experience for Hindi-speaking users. Tasks: Rated results for intent match, relevance, linguistic accuracy, and webpage quality; assessed ads and landing pages; flagged misleading or low-quality content; provided structured feedback to improve rankings. Project Size: Processed a high volume of search queries across diverse topics and user intents. Quality Measures: Followed strict rating guidelines, ensured cultural and language accuracy, maintained consistency, and adhered to confidentiality standards.

2024 - 2025

Education

Work History