Keychain | Lead Backend Engineer | Sept 2026 | Rejected
Summary
I interviewed for a Lead Backend Engineer role at Keychain in September 2026. I completed an exploratory call, an HLD design round, a take‑home assignment, an assignment discussion, and a final CTO round, but was ultimately rejected.
Full Experience
Current Org: Small Startup College: Tier-1 YOE: 8+ Current Comp: 60L Base + 5L Bonus + ESOPS
Exploratory Call With Senior EM. Normal behavioural and project discussions. Company and team intro, your introduction, why switch.
Round 1 - HLD Design a job scheduler that runs jobs at a specific time in the future. Job can run on a 10 second delay max.
Take Home Assignment Make a distributed priority queue. Note: You can use a simple single node in-memory queue here. I over-engineered it and made a queue with a gateway and actual distribution among various nodes and got confused in concurrency handling in the assignment discussion.
Round 2 - Assignment discussion Walk through of your code.
- Focus more on concurrency, locking, how will you scale, max throughput.
- Measure max throughput beforehand only.
- Structure your code properly and know how things are handled in detail.
Round 3 - With CTO Intro discussion and previous projects. Some technical questions:
- Why we chose MongoDB in our use case, why not Postgres.
- How Mongo Atlas manages AWS to GCP transfer with zero downtime.
- If you have two microservices that give say user and order data, now you want to combine data, how will you do it.
- If you have two microservices that give data and you want to combine data and provide filtration over combined data, how to do it?
Verdict: Rejected. No feedback shared.
Interview Questions (6)
Job Scheduler Design
Design a job scheduler that runs jobs at a specific time in the future. Job can run on a 10 second delay max.
Distributed Priority Queue
Make a distributed priority queue. Note: You can use a simple single node in-memory queue here. I over-engineered it and made a queue with a gateway and actual distribution among various nodes and got confused in concurrency handling in the assignment discussion.
Choice of MongoDB vs PostgreSQL
Why we chose MongoDB in our use case, why not Postgres.
Mongo Atlas Cross-Cloud Data Transfer
How Mongo Atlas manages AWS to GCP transfer with zero downtime.
Combining Data from Two Microservices
If you have two microservices that give say user and order data, now you want to combine data, how will you do it.
Aggregating and Filtering Data from Multiple Microservices
If you have two microservices that give data and you want to combine data and provide filtration over combined data, how to do it?