remove wiki

fix wikipedia search url

add tools support

changing response way llm

changing response way

Update system model to Gemma 3 1B Instruct and humanize responses

fix: improve question matcher keyword normalization, expanded stop words, and use full XML context in LLM fallback

feat: add LLM-based query restructuring and grammar correction stage

feat: implement semantic chunking, dual-retrieval cross-lingual search, and direct Q&A bypass routing

Fix Hindi/Hinglish language output alignment using retrieval translation pipeline

fix: add threading lock synchronization to prevent concurrent model inference crashes

feat: optimize response time by offloading translation to deep-translator and reducing ctx/retrieval k size

data: update source documents to hindi_interview_questions_fixed.pdf

chore: revert embedding model to all-MiniLM-L6-v2 since query translation handles English retrieval

feat: implement multilingual query translation and language-specific prompting

feat: return custom sorry message if context is missing or not provided

feat: implement multilingual RAG pipeline for English, Hindi, and Hinglish

Optimize Hugging Face Space: add eager model loading, reduce max tokens, fix stop tokens, limit CPU threads

feat: implement complete RAG-based AI engine with Ollama fallback and vector search support

Optimize startup speed and query latency

add new documents

add README.md with Hugging Face Space metadata

Initial clean release