PracHub
QuestionsPremiumLearningGuidesCheatsheetNEWCoaches
|Home/ML System Design/Google

Design a chatbot over structured and unstructured data

Last updated: May 7, 2026

Quick Overview

This question evaluates a machine learning engineer's ability to design end-to-end systems that integrate structured and unstructured data, testing competencies in data architecture, information retrieval, LLM orchestration, access control, privacy, and monitoring within the ML system design domain.

  • medium
  • Google
  • ML System Design
  • Machine Learning Engineer

Design a chatbot over structured and unstructured data

Company: Google

Role: Machine Learning Engineer

Category: ML System Design

Difficulty: medium

Interview Round: Onsite

Design a chatbot that can answer user questions using both: - **Structured data** (e.g., relational tables such as orders, products, pricing, user accounts), and - **Unstructured data** (e.g., documents, knowledge base articles, FAQs, PDFs, internal wikis). The chatbot should: - Provide accurate, grounded answers - Support follow-up questions (multi-turn) - Cite sources when possible - Respect permissions (users can only access authorized data) Discuss: 1. High-level architecture and main components 2. Data ingestion and indexing strategy for both data types 3. Retrieval strategy (including when to query SQL vs retrieve documents) 4. LLM prompting / tool-calling approach 5. Safety, privacy, and access control 6. Evaluation (offline + online) and monitoring 7. Latency/cost considerations and scalability

Quick Answer: This question evaluates a machine learning engineer's ability to design end-to-end systems that integrate structured and unstructured data, testing competencies in data architecture, information retrieval, LLM orchestration, access control, privacy, and monitoring within the ML system design domain.

Related Interview Questions

  • Design an app-store app recommendation system - Google (medium)
  • Design a fraud detection system - Google (medium)
  • Choose Fast or Cheap Models - Google
  • Design ML system for self-driving perception - Google (medium)
  • Design a real-time recommendation system - Google (hard)
Google logo
Google
Feb 8, 2026, 12:00 AM
Machine Learning Engineer
Onsite
ML System Design
7
0
Loading...

Design a chatbot that can answer user questions using both:

  • Structured data (e.g., relational tables such as orders, products, pricing, user accounts), and
  • Unstructured data (e.g., documents, knowledge base articles, FAQs, PDFs, internal wikis).

The chatbot should:

  • Provide accurate, grounded answers
  • Support follow-up questions (multi-turn)
  • Cite sources when possible
  • Respect permissions (users can only access authorized data)

Discuss:

  1. High-level architecture and main components
  2. Data ingestion and indexing strategy for both data types
  3. Retrieval strategy (including when to query SQL vs retrieve documents)
  4. LLM prompting / tool-calling approach
  5. Safety, privacy, and access control
  6. Evaluation (offline + online) and monitoring
  7. Latency/cost considerations and scalability

Solution

Show

Comments (0)

Sign in to leave a comment

Loading comments...

Browse More Questions

More ML System Design•More Google•More Machine Learning Engineer•Google Machine Learning Engineer•Google ML System Design•Machine Learning Engineer ML System Design
PracHub

Master your tech interviews with 7,500+ real questions from top companies.

Product

  • Questions
  • Learning Tracks
  • Interview Guides
  • Resources
  • Premium
  • For Universities
  • Student Access

Browse

  • By Company
  • By Role
  • By Category
  • Topic Hubs
  • SQL Questions
  • Compare Platforms
  • Discord Community

Support

  • support@prachub.com
  • (916) 541-4762

Legal

  • Privacy Policy
  • Terms of Service
  • About Us

© 2026 PracHub. All rights reserved.