Ritesh Khatarkar

Generated targeted voice data to train conversational AI systems with an emphasis on tonal accuracy, clarity, and natural pacing. Maintained linguistic quality and followed recording requirements for effective training data.

Transcribed complex audio and video files with high accuracy for AI training purposes. Captured nuanced speech patterns and maintained transcription quality aligned to project requirements.

Accurately categorized and labeled complex datasets to improve machine learning and computer vision model performance. Ensured high-quality outputs by meticulously following detailed annotation frameworks and project protocols.

Evaluated and refined AI-generated responses to enhance the natural language capabilities of advanced models. Performed diverse linguistic tasks using bilingual knowledge to assess contextual accuracy while adhering to project guidelines.

Optimized large language model responses through RLHF by ranking outputs for accuracy, safety, and logical flow. Verified model outputs and authored high-complexity prompts and gold-standard responses to improve instruction following and factual reliability.

Generated authentic remote Hindi audio datasets by engaging in real-time conversations with strangers for AI model training. Produced highly accurate Hindi audio transcriptions to support text-to-speech and speech-to-text model refinement while following strict technical guidelines.

Recorded and annotated high-fidelity Hindi and English audio datasets for proprietary speech recognition and synthesis model training. Ensured consistent labeling quality to support reliable model development.

Bachelor of Engineering in Computer Science from RGPV, completed in 2010.

Completed Higher Secondary with a focus on Mathematics at Satpura Higher Secondary School in 2006.