DSA questions
- compress string (aaabbc --> a3b2c1)
- char frequency in string
- perform left rotation on a list by d, without using additional list
Gen AI questions
- Talk about recent gen ai project
- Detailed question about RAG architecture ( pre processing, chunking, embedding, vector db)
- How exactly storage and retrieval happens in vector db
- Type of vector db and when to use which one
- in memory vs persistent vector db
- semantic search in detail, formula of cosin similarly
- How to handle caching, if user uploads same file in a chatbot, how to handle
- HNSW, IVF, what is the use and their working
- HNSW available in which db
- how you pass chat history to llm
- how is chat history passed to llm in langchain
- short term and long term memory, how to implement in langchain
- how to build custom tools, in langchain / dspy/ crewai
- tool decorator
- how tool decorator uses docstring of tool function
- Under the hood how tool decorator works
- how to pass info like session ID to tool, and also not expose it to llm
- how to pass tools to an agent
- if a value is not passed by llm how to get that in the tool, will it fail?
- LangGraph basics
- Pydantic and its uses
- what llms have you worked with (list all models / versions, eg: gpt 4 --> gpt 4.1 , 4o, 4 mini etc, gemini--> 1.5 flash 1.5 pro etc)
- open source models
- hugging face vs ollama
- How to handle a CSV file in RAG
- how to handle the table data (unstructured.io)
- have you tried mistral for extracting table?
- pypdf, pdfplumber and what other libraries have you tried for table extraction, which works best?
- how to make image embeddings, what model is used
- multimodal llms (take image/ audio/video/ file as input, and gives multimodal output)
- Talk about any other project of yours
- Few questions related to project
- Have you also worked on ML models apart from gen ai?
- How do you evaluate RAG?
- What is precision and recall in RAGAS, also in ML, formula and definition
- Any questions for me..