Praveen M.

About Praveen M.

AI/ML Data Annotation | LLM/NLP Dataset Development | PII Detection | Endpoint Security & Malware Troubleshooting. I’m a detail-oriented and analytical professional with 6+ years of hands-on experience in AI/ML data annotation, LLM dataset development, and PII/data privacy model support. I specialize in building high-quality, human-validated datasets that strengthen NLP, NER, and Large Language Models helping organizations deploy AI responsibly and at scale across my career with Amazon, HCLTech (PayPal project), ACL Digital, and Concentrix, I’ve worked on multimodal annotation (text, audio & image), sensitive context detection, dataset indexing, SQL-based data extraction, and model fine-tuning support. I enjoy solving ambiguity through analysis, improving model accuracy with error-pattern insights, and contributing to smooth ML operations with structured QA and documentation.Along with AI/ML experience, I also bring a strong technical foundation in endpoint protection and enterprise support, including malware investigation, OS troubleshooting, SLA ticket resolution, and customer handling.I’m passionate about advancing ethical and responsible in optimizing ML pipelines through quality datasets. learning new tools and adapting to dynamic work environments I thrive in teams where accuracy, ownership, and collaboration matter. Always open to learning, new opportunities, and meaningful conversations around AI, automation, and secure digital transformation.

Experience

1. Lead Engineer

   HCLTech

   May 2023 — Present · Chennai, IN

   Working for client ( PAYPAL ),Spearheaded the end-to-end data annotation, curation, and fine-tuning pipeline for PayPal’s Data Privacy (PII Detection) models, directly improving model precision and recall.Fine-tuned contextual models using large-scale real-world UDC (User Data Corpus) to ensure high reliability of PII detection in production environments.Performed high-accuracy manual labeling and QA, acting as the Ground Truth (GT) expert for identifying sensitive entities across 20+ PII categories in compliance with PayPal guidelines.Identified and analyzed complex prediction scenarios (TN, FN, FP, TP) to generate targeted training datasets, reducing model bias and error patterns.Delivered sensitive context models with 97% verified accuracy and high confidence scores, supporting secure and reliable model deployment.Led data indexing for LLM training, preparing optimized and scalable datasets for generative AI applications.Enhanced NER, InfoSec, and Sensitive Class Classifier models, including improved language detection and person-aware contextual understanding (1st / 2nd / 3rd-person perspectives).Developed complex SQL queries to extract and assemble training datasets from internal proprietary sources and external public repositories.Applied synthetic text generation and strategic data curation to enrich low-resource sensitive categories, ensuring balanced dataset representation for ML training.

Education

• SBIOA Public School - India

  higher secondary certificate, Computer Science

• Dr MGR Educational and Research Institute

  Bachelor's degree, Mechanical Engineering

• Kendriya Vidyalaya

  SSLC, Computer Science