Sanskrit, Kannada, and Hindi, as morphologically rich Indian languages, present a significant challenge in NLP due to their complexity and the limited available resources. A critical step in many NLP processes is POS tagging, where each word in a sentence is automatically categorized based on its grammatical function. This work introduces a Hybrid POS tagger specifically designed for Sanskrit, Kannada, and Hindi. The system combines Conditional Random Field (CRF) with the Hidden Markov Model (HMM), allowing it to handle linguistic features that are beyond the capabilities of a traditional HMM-based POS tagger. The assessment of the Hybrid POS tagger has shown high accuracies across the three languages, thus proving to be a reliable tagger for higher level NLP tasks.