Lyft Engineering Manager Interview Questions

1.Tell me about a major production incident your team owned end-to-end. Walk me through how you led the response — not just the technical fix, but the people and process side.

2.Tell me about a time your team's on-call rotation was broken — engineers were burning out, alerts were noisy, or incidents kept recurring. How did you fix it, and how long did it actually take?

3.Describe a time your team was building something with a hard real-time constraint — matching, routing, ETAs, anything time-sensitive. What technical bets did you push your team to make, and how did they play out?

4.Walk me through a time you took part in a design review for a system outside your team's scope — and your feedback actually changed the direction of the design. What did you catch, and how did you make your case?

5.Tell me about a time you identified a safety or trust risk in a system your team owned or was building. How did you make the call to prioritize it, and who did you have to convince?

6.Describe a time you pushed your team to break apart a monolith or consolidate services that had fragmented too far. What drove the decision, and what was the hardest part of the migration?

7.Your team owns a Kafka consumer that processes driver location updates — it's falling behind during surge periods. How do you think about diagnosing and fixing consumer lag at this scale?

8.How do you think about managing technical debt across a team that's shipping features quickly? What's your actual process for keeping it from compounding?

9.Walk me through how you'd design the data model and write path for Lyft's driver earnings ledger — the system that tracks every trip payout, bonus, and deduction for a driver's pay statement.

10.Your team is adding a new service to the request path for ride matching — it needs to call an ML model to score driver candidates before dispatch. The matching system currently has a 200ms P99 latency budget. How do you think about fitting this in?

11.You're inheriting a team whose core service is a Python monolith that handles both trip pricing logic and driver incentive calculation — they're deeply entangled. Leadership wants pricing to move faster but the incentives team is blocked on every deploy. How do you approach the decomposition?

12.Your team is building the backend for a new feature that lets riders see real-time vehicle position on the map during an active ride — updated every second. Walk me through the architecture from the driver's phone to the rider's screen.

13.Walk me through how you'd design the configuration system that controls Lyft's surge pricing multipliers — operators need to push new pricing caps to all regions within seconds, but a bad config push could corrupt pricing for millions of active rides.

14.Design the system that tracks real-time dock availability for Lyft's bike-share stations — Citi Bike in NYC alone has over 1,800 stations, and riders need to know how many bikes and open docks are available before they walk to a station.

15.Design the backend for Lyft Business's ride-ordering portal — a corporate customer like a hospital needs to book rides on behalf of patients, enforce spend policies (e.g., no rides over $50), and get a monthly invoice. Walk me through the architecture.

16.Design the ETAs system that shows a rider 'your driver arrives in 4 minutes' — it needs to be accurate enough that riders trust it, and it updates live as the driver moves. What does the architecture look like, and where does it break?

17.Design the fraud detection system that decides, in real time, whether to allow a rider to complete a ride payment — Lyft processes millions of transactions per day and can't afford false positives that block legitimate riders, but chargebacks from fraudulent rides cost real money. How do you build it?

18.Design the system that powers 'driver mode' on the Lyft Driver app — specifically the state machine that tracks whether a driver is offline, online-waiting, matched, en-route to pickup, and in-trip — at a scale of hundreds of thousands of concurrent active drivers.

19.Tell me about a time you had to deliver critical feedback to a senior engineer on your team — someone technically strong but whose behavior was creating friction for others. How did you handle it?

20.Describe a time you had to tell a senior engineer their project was being killed or descoped significantly — work they'd been heads-down on for months. How did you communicate it and help them land on the other side?

21.Tell me about a time you had to align two engineering teams — maybe yours and a peer's — that had competing priorities or genuinely different opinions on how something should be built. How did you get to a decision?

22.You're hiring for a senior engineer role on your team and your panel gives mixed feedback — two interviewers say strong hire, two say no-hire, and the deciding vote is yours. What's your actual process for making that call?

23.Tell me about a time you owned a platform or service that multiple other teams depended on. How did you manage competing feature requests from those teams while keeping your own team's roadmap coherent?

24.Describe a time you had to kill a technical direction your team had been advocating for — because external reality changed, or you got new information. How did you reverse course without destroying the team's confidence in decision-making?

25.Estimate how many driver-hours Lyft loses per day to rides that get canceled after a driver is already en route to pickup. What levers would you pull to reduce that?

26.Lyft's average ETA accuracy is measured as the delta between predicted and actual pickup time. If that number degrades by 30 seconds on average across all markets starting Monday morning, how would you figure out what happened?

27.Lyft's bikeshare ridership — Citi Bike, Bay Wheels — drops about 60% in winter months. How would you estimate the fixed infrastructure cost per ride that Lyft absorbs during that off-peak period, and how should that inform product strategy?

28.Your team owns Lyft's promo code redemption system. You notice that promo redemption revenue impact is 15% higher than your discount budget models predicted for the past two weeks. No new promos launched. How do you think through this?

29.Estimate the incremental infrastructure cost Lyft would incur if it expanded its Lyft Business healthcare rides program — hospital-to-patient transport — from 50 cities to all 300+ markets where Lyft operates. What engineering investments drive the majority of that cost, and where's the leverage?

30.Lyft's ride acceptance rate among drivers drops 12% during a city-wide event — say, a stadium concert in a market you own. The ops team is asking engineering to 'fix it.' How do you structure the problem and decide what, if anything, your team should build?