1. Introduction
2. How many hashtag#genAI projects have you done so far? Tell about tools, technologies and cloud infra you have worked with for these projects.
3. What is regularization? -> Explain L1 and L2 and how these are different from one another?
4. What is a cost function? why it is needed? give examples.
5. Bias-Variance tradeoff
6. Explain various fetaure selection techniques ?
7. What do you mean by ensemble techniques? - > How boosting is different from bagging ? -> How bootstrap and aggregation is achieved for bagging techniques?
8. Explain working if random forest. -> Different hyper parameters for RF.
9. How do you detect and handle outliers?
10. WHat is postitional encoding ? -> why is this needed? ->What are different techniques for this? -> why this was not needed for RNN/LSTM?
11. Explain hashtag#transformer architecture -> explain different decoding techniques?
12. Explain hashtag#RAG architecture
13. Explain different chunking strategies
14. what is semnatic based chunking?
15. What is semenatic search?
16. Why cant we store embeddings in a conventional data base store?
17. hashtag#Python question : Find the pairs on indices from a sorted list for which sum of elements is equal to a given target.
e.g. li=[1,2,3,4,4,6,7,9], target= 8 the output= [[0,6],[1,5],[3,4]]
follow up -> suggest a solution with O(n) time complexity.
Round - 2
1. Introduction
2. Explain the complete CI/CD process for your hashtag#GCP project. -> covered versioning control, docker, Jenkins, Secret manager, GKE.
3. How did you create workloads and other resources required for hashtag#deployment?
4. What are your roles and responsibilities in your current organization?
5. Have you created hashtag#docker image from scratch? What are some important components or configurations included in a hashtag#Dockerfile ?
6. How do you deploy a new version of code to hashtag#GKE across dev/QA/prod environment ?
7. How did you monitor deployment logs and troubleshoot deployment issues?
8. How do you configure hashtag#ingress and routing paths for your application?
9. What all services have you used from hashtag#vertexAI, give a brief of each.
10. Explain architecture of one of your GCP project along with tech stack.
11. How sub-agents are created and invoked using hashtag#google hashtag#adk.
12. How did you evaluate your hashtag#NLP2SQL use case?
13. Where were your hashtag#agents deployed? How did you add hashtag#observability to your hashtag#agentic solution?
14. How did you setup hashtag#RAGEngine in hashtag#vertexAI ?
15. Why didn’t you use Vertex AI’s GenAI observability, traceability, and evaluation tracking features initially? If required, how would you integrate them into your application now?