Indexing Internals - Qdrant | CodeMentor | NextRound.in

Technologies Coding Problems Bookmarks Learning Paths Login

Technologies Coding Problems Bookmarks Learning Paths Login

</>

Loading Topics...

AI-powered interview preparation platform. Practice with curated questions, mock interviews, and personalized learning paths to crack your dream tech interview.

Quick Links

Technologies
Mock Interviews
Saved Questions
Pricing

Company

About Us
Contact Us

Legal

Privacy Policy
Terms of Use

© 2026 nextRound. All rights reserved.

1What does HNSW stand for, and at a high level, how does it achieve sub-linear approximate nearest-neighbor search?ExpertHigh 2What do the HNSW parameters m and ef_construct control, and what trade-off do they represent?ExpertHigh 3What does the query-time parameter ef (search breadth) control, and how would you use it to trade off recall against latency?ExpertHigh 4Why might increasing m significantly improve recall on one dataset but barely help - or even hurt latency - on another?ExpertHigh 5Why does Qdrant set m: 0 on a named vector used purely for reranking (e.g. a ColBERT multivector)?ExpertHigh 6What problem does vector quantization solve, and what is the fundamental trade-off it introduces?ExpertHigh 7Compare scalar quantization, product quantization, and binary quantization in Qdrant in terms of compression ratio and accuracy impact.ExpertHigh 8What are oversampling and rescoring in the context of binary quantization, and why are they necessary?ExpertHigh 9What newer quantization options - beyond the original scalar, product, and binary trio - has Qdrant introduced to fine-tune the compression/accuracy curve?ExpertHigh 10What is Inline Storage, and how does embedding quantized vectors directly into HNSW graph nodes improve disk-based search performance?ExpertHigh 11What is a multivector point, and how does it differ from a point with several named vectors?ExpertHigh 12How does late-interaction scoring (as used by ColBERT-style models) with MaxSim differ from comparing two single dense vectors?ExpertHigh 13Why is late-interaction reranking typically applied to a small candidate set rather than the entire collection?ExpertHigh 14Design a three-stage retrieval pipeline using dense retrieval, sparse retrieval, fusion, and ColBERT reranking. What does each stage contribute?ExpertHigh

All Topics

1

Setup and Basic Use

2

Named Vectors and Multitenancy

Sparse Vectors and Hybrid Search

3

Snapshots and Backups

Indexing Internals

Internal Architecture

4

Performance and Optimisation

System Design and Scalability

5

Error Handling and Debugging

Design Patterns

Implementation Scenarios