Slack Engineering Manager Interview Questions

1.Think of a time you noticed a quieter engineer on your team had strong convictions about a technical direction but wasn't getting airtime. What did you do, and how did it play out?

2.Tell me about a code review comment you wrote that you later realized was wrong — either technically or in how you delivered it. What happened?

3.Walk me through a time when customer feedback — whether from support tickets, a user interview, or usage data — caused you to reprioritize something your team was already mid-sprint on. What did you learn, and was it the right call?

4.Tell me about the most heated disagreement you've had with a PM or design partner in the last two years. What was at stake, and how did it get resolved?

5.Has there ever been a time you changed the structure of a meeting or a written process specifically to surface input from engineers who weren't naturally volunteering it? What did you change, and did it work?

6.Tell me about a time you intervened in your team's code review culture — either because the reviews were becoming too harsh, too rubber-stampy, or inconsistently applied across engineers. What did you change and how?

7.Your team owns a service that publishes messages to Kafka and downstream consumers are falling behind. How do you diagnose whether the problem is the producer, the broker, or the consumers — and how do you lead your team through that investigation?

8.Your team is migrating a high-traffic MySQL table from one shard topology to a new one — live, without downtime. What's your role as the EM in that migration, and what does a safe execution plan look like?

9.Slack Connect lets companies share channels across organizational boundaries. If you were leading the team that owns the messaging delivery path for Connect channels, what are the two or three hardest technical problems you'd expect to own, and how would you staff and prioritize against them?

10.Your team maintains a Go service that's part of Slack's real-time message delivery path. A new Workflow Builder feature requires this service to call out to a third-party webhook synchronously in the hot path. How do you advise your team on that design, and what's the conversation you have with the PM?

11.Slack's Electron desktop client and the iOS and Android apps all need to show consistent channel state — unread counts, message ordering, thread replies — across devices. Walk me through how you'd architect the client-side state synchronization layer, and what failure modes keep you up at night.

12.Agentforce in Slack routes user queries to AI agents that may call Salesforce APIs, Slack's own search index, and third-party tools in a single response. If you were leading the team that owns that orchestration layer, how would you design it for reliability and latency — and how do you make the SLO story work when you're dependent on four external systems?

13.Walk me through how you'd design the read-state tracking system for Slack — the thing that knows which messages a user has and hasn't seen across every channel and DM.

14.Slack Workflow Builder lets non-developers publish workflows that trigger on channel events — a new message, a reaction, a form submission. How would you design the event routing layer that gets a channel event to the right set of subscribed workflows reliably?

15.Design the presence system — the green dot, away indicator, and 'active on mobile' badge — for a workspace platform used by millions of concurrent users. How does it stay accurate without melting Slack's infrastructure?

16.Slack Huddles need to let a user spontaneously start an audio call in a channel and have teammates join within seconds. Walk me through how you'd design the signaling and session management layer — not the media transport itself, just the infrastructure that coordinates who's in, who's speaking, and channel membership.

17.Design a search indexing pipeline for Slack messages that can handle billions of messages, near-real-time indexability, and per-message access control — where a user should only see messages from channels they're currently a member of.

18.Design the rate limiting and quota enforcement layer for Slack's platform API — the one that thousands of third-party apps use to post messages, read channel history, and call workspace APIs. How do you enforce per-app and per-workspace limits without those limits becoming a single point of failure?

19.Tell me about a time you had to coach an engineer on how to give better code review feedback — not the technical content, but the tone or delivery. What did you actually say to them?

20.Think about the best working relationship you've built with a designer on a product team. What made it work, and what would have broken it?

21.Tell me about a time you had to advocate for a major technical investment — refactoring, infrastructure migration, paying down debt — to a skeptical product organization. How did you make the case, and did it land?

22.Describe a time you had to deliver critical feedback to a senior engineer — someone more tenured than most of your team — who was a respected technical voice but was creating friction on the team. How did you approach it?

23.Tell me about a time your team shipped something that customers were technically using but clearly wasn't meeting their actual need — not a bug, but a misunderstanding of the problem. How did you find out, and what did you do?

24.You're six months into managing a team and you realize the team's velocity is strong, the engineers are happy, but the team is becoming increasingly isolated — shipping well in a silo, but drifting out of alignment with where the broader engineering organization is headed. How do you recognize that pattern and what do you do about it?

25.Slack's daily active user count is in the tens of millions. Estimate how many WebSocket connections Slack's real-time messaging servers are holding open at any given moment during peak business hours.

26.You're reviewing your team's quarterly roadmap and you realize 60% of the planned work is feature work and 40% is reliability and infrastructure. A VP is pushing to flip that to 80/20 in favor of features for the next two quarters. How do you think through whether to push back?

27.Slack's message notification delivery rate drops from 99.5% to 97% over a two-week period. No incident was filed, no major deploy went out. How do you structure your investigation, and what hypotheses do you prioritize first?

28.Slack is considering charging enterprise customers for Slack Connect usage based on the number of external organizations they're connected to. As an EM whose team owns the Connect messaging infrastructure, what are the two or three largest technical bets you'd need to make to actually support that business model?

29.Slack's engineering organization is growing. Estimate what percentage of total engineering cost is consumed by on-call, incident response, and post-mortem work in a typical quarter — and tell me how you'd use that number as an EM to make decisions.

30.Agentforce in Slack is embedding AI agents into the core messaging experience. If AI-generated messages in channels become common, walk me through how the metrics you'd use to evaluate your team's success would need to evolve — and what new failure modes you'd be watching for that don't exist in a purely human-to-human messaging product.