Software Engineer · Onsite · Multiple Types

Stripe — Software Engineer ✅ Passed

Level: Mid-Level

Round: Onsite · Type: Multiple Types · Difficulty: 7/10 · Duration: 240 min · Interviewer: Unfriendly

Topics: Object-Oriented Programming, Data Structures, Algorithms, API Integration, Debugging, Behavioral Questions

Location: San Francisco Bay Area

Interview date: 2026-01-15

Summary

Interview Rounds Overview

• Round 1: Technical (Email Subscription)
• Round 2: Integration (Bikemap API)
• Round 3: Debugging (Mako Python)
• Round 4: Behavioral

Details

Technical: Email Subscription

I was given the following inputs:

• user_accounts: A list of users, each with a name, plan, duration, and account date.
• schedule: A dictionary with keys for start, end, and a negative number representing the number of days before plan expiration to send an upcoming expiry email. Values are email titles (e.g., Welcome, Upcoming Expiry, Expired).
• subscription_change: A list containing change_time, name, new_plan, or extension_days.

I needed to generate outputs like:

0: [Welcome] Subscription for John (Silver) where 0 is the time, Welcome is the email title, John is the username, and Silver is the user's plan.

Tasks:

1. Given user_accounts and schedule, print emails in chronological order.
2. Handle subscription_changes. On the day of a plan change, send an 'Updated' email. Subsequent emails should reflect the new plan.
3. Incorporate extension in subscription_change. On the day of extension, send a 'Renewed' email. Update upcoming expiry and expired email times accordingly.

Follow-up:

• How would you test this?

Integration: Bikemap API

I had to work with the Bikemap API to:

1. Read geometric data from a JSON file.
2. Make API requests and download images, focusing on image storage.
3. Make API requests to annotate points (read from task 1) on an image and connect them with lines.
4. Read landmark data from another JSON file and make API requests to annotate landmarks, finding the nearest point for each landmark.
5. (Not required) Calculate speed, deemed too complex.

Debug: Mako (Python)

I chose Python with the Mako template engine. I encountered these issues:

1. Missing a check to differentiate between a folder and a file.
2. Missing a visit function, causing arguments not to be written.
3. (Not required by the interviewer) Missing a check to ensure ControlLine only contains comments.

Behavioral Questions:

These were standard interview questions.

• How to deal with conflicts?
• How you work across teams?
• Most impactful project?
• Experience mentoring others?
• How would your current manager describe you?