← Google Interview Insights

Google·Technical Product Manager·Onsite - System Design / Architecture·Staff

Staff
May 2026

Summary

Google TPM system design round, one big open-ended question about mapping infrastructure at global scale. Felt like a logistics puzzle dressed up as a product problem, and I kept second-guessing whether they wanted me to go deeper on the tech or the operational side.

Questions Asked (1)

Q1

You have a fleet of 1,000 cars to photograph every address on Earth for a Street View-style product. How do you design the system end to end?

System DesignTechnical Trade-offsProduct Strategy
Author's notes

Spent the first few minutes thinking about routing and coverage, which felt right, but I underweighted the data pipeline side.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the product goals and constraints, then break the problem into major components: data acquisition, processing, storage, and serving. Discuss trade-offs between coverage, quality, cost, and latency, and propose a phased rollout with metrics for success.

Pro tip: Emphasize the importance of prioritization: not all addresses are equally valuable, so focus on high-density areas first and use a tiered approach to balance cost and coverage. Also, mention the need for a feedback loop to continuously improve data quality.

1. Clarify Requirements and Constraints

Ask questions to understand the scope: What resolution? How often updated? What's the budget? Are there privacy concerns? This ensures alignment with product goals.

2. Design Data Collection Strategy

Plan how to deploy 1,000 cars efficiently: route optimization, scheduling, sensor setup (cameras, GPS, LiDAR), and data upload mechanisms. Consider partnerships or crowdsourcing for hard-to-reach areas.

3. Build Data Processing Pipeline

Outline steps for ingesting, processing (stitching images, blurring faces/license plates), and storing data at scale. Discuss use of cloud services, batch vs. stream processing, and quality checks.

4. Design Serving and Access Layer

Define how users will access the imagery (APIs, web app) and how to serve low-latency queries. Consider caching, CDN, and indexing by address/geolocation.

5. Define Metrics and Iterate

Establish success metrics (coverage %, freshness, user engagement) and a feedback loop for continuous improvement. Plan for scaling and maintenance.

Key Points to Mention

  • Prioritization based on population density and business value
  • Scalable data processing using cloud services (e.g., GCP) and parallelization
  • Privacy and legal compliance (blurring, consent, local regulations)
  • Cost optimization: trade-offs between image quality, frequency, and storage
  • Use of machine learning for automated quality control and object detection
  • Phased rollout with MVP in select cities, then expand

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.