J'ai postulé via un recruteur. Le processus a pris 3 mois. J'ai passé un entretien chez Instacart (San Francisco, CA) en juil. 2017
Entretien
Approached by recruiter through email. Followed by a 15 minute hackerrank challenge. Then talk to either a lead or manager for 45 minutes. Followed by a take home web app project to be done in 4 hours. After which you go for an onsite. First one is a pair programming challenge followed by 3 to 4 behavioral interviews and one lunch interview. Finish with talking to HR/recruiter.
I did everything well but they never got back to me with the decision. Waited for four weeks and they had no clue what the hiring managers decision was. Waste of time.
Questions d'entretien [3]
Question 1
PubSub implementation in Hackerrank. Very straightforward.
Pair programming: Hashmap with time awareness i.e. Key can store multiple values based on the time it was put. When getting can provide no timestamp (get last inserted value), or one matching either the timestamp or the closest value to the given timestamp.
I'm sure it's easier to grade applicants with AI, but it'd be nice to have humans involved in the interview process, too. You'll face a couple of AI powered coding challenges. The bigger problems are the ethical implications here after looking into how the company works, with variable pricing based on data on how badly they can screw the consumer with upcharges and how to keep money in their pockets rather than the shoppers & drivers
Questions d'entretien [1]
Question 1
If you could replace your family with AI, would you do it?
J'ai postulé en ligne. J'ai passé un entretien chez Instacart (San Francisco, CA) en févr. 2026
Entretien
First there is a codesignal test. This is for 90 minutes and felt frantic. Way too many features/code paths to implement.
Then there is the karat llm interview which is 50 minutes.
Then a full day of regular interview loop.
Questions d'entretien [1]
Question 1
For CodeSignal it was implementing a cache key store with timestamps. Started with basic setting and getting, then evolved into more advanced code. 4 different parts. Writing a lot of code for this.
For the LLM interview you work with a company called Karat. You are giving an VS code like workspace and an ai to help. You have 50 minutes. They then dump a large project in your space. For me it was a bus route simulator.
I was then asked to implement a feature : Have the system prioritize Priority Pass people. I used the LLM to track down the code and implement it, but it was difficult to test in code that is unknown to me. The next feature was implementing wheelchairs for the bus which requires special capacity checks.
J'ai postulé en ligne. Le processus a pris 4 semaines. J'ai passé un entretien chez Instacart (Toronto, ON) en févr. 2025
Entretien
1. Online OA with codesignal to start with. Very standard and not hard
2. Got the 4 interview rounds and you can always divide into 2 parts of 2 interviews each. It was be 2 codesignals with medium difficulty and I managed to do both. One system design which was average for me and interviewer was okay as well. Behavioural round was standard. No wonder why I didn't manage to get the offer.