Summary
Paras Dahiya is a Senior Data Scientist with 8 years of experience building production-ready ML and computer vision systems, currently automating clinical prior-authorization workflows at Carelon Health using retriever-augmented generation and large language models. His background spans industry projects from self- and semi-supervised tire-defect classification and form understanding for Continental to real-time road segmentation and pothole detection for IBM, and multiple in-house OCR/handwritten-text recognition systems. Comfortable with unstructured data and class-imbalanced problems, he combines research-oriented experimentation with practical engineering to deliver high-impact automation. Based in Bengaluru, he pairs strong academic foundations from IIIT SriCity with hands-on experience across startups and enterprise teams, and has a knack for turning noisy, low-quality inputs into reliable structured outputs. An underappreciated strength is his repeated role building custom OCR pipelines—showing deep expertise in bridging vision, heuristics, and domain-specific preprocessing.
8 years of coding experience
5 years of employment as a software developer
Sciences, 94% at intermediate level and 10.0 CGPA at matriculation, Sciences, 94% at intermediate level and 10.0 CGPA at matriculation at Little Angels Senior Secondary School, Sonipat
Bachelor of Technology, Computer Science, 8.92, Bachelor of Technology, Computer Science, 8.92 at Indian Institute of Information Technology, SriCity
English, Hindi