← Back to Directory

Salesforce interview notes, page 12 of 13

Large Enterprises

Most of these Salesforce write-ups come from Software Engineers (104), then Product Managers (6) and Backend Engineers (4). The most reported rounds are the technical phone screen (65), the system design round (31), the hiring manager screen (6) and the behavioral / leadership round (5).

121 interview notes · updated Sep 2026

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce MTS coding interview with two algorithm problems. Sailed through the first one but completely hit a wall on the second, and now I'm just waiting to see if partial credit is a thing.
  • Find all duplicate elements in a given array.
  • Given a directed tree, find the minimum number of edge reversals needed so that every node is reachable from any given root node.

“Nailed this one, follow-ups included.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce SWE interview with a string algorithm problem that looks deceptively simple until you remember the cyclic part.
  • Given a cyclic string S and a pattern k, find the shortest contiguous subsequence of S (treating it as circular) that contains all characters in k.

“The example they gave was s = 'abcdefgh' and k = 'fha', answer being 'fgha'.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce software engineer interview with a dynamic programming problem that looks straightforward until you actually sit down and think about it. The problem was well-defined but the optimal approach took me longer to see than I'd like to admit.
  • Given an array of campaign costs and a number of weeks, partition the array into exactly that many contiguous non-empty groups. Each week's cost is the max of its group. Return the minimum possible sum of those per-week maxima.

“My first instinct was greedy and it was wrong.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Interviewed for a solutions architect role at Salesforce and got hit with a networking fundamentals question I wasn't expecting to go as deep on as it did.
  • How do CDNs work?

“Thought I had this cold but partway through my answer I realized I was just describing caching at a high level and hadn't touched on edge nodes, origin pull vs push, or how routing decisions actually get made.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce software engineering interview with a single algorithmic problem. Pretty focused session, just one coding question the whole time.
  • Given a positive integer n, you can add or subtract any power of 2 from n in each operation. What is the minimum number of operations needed to reduce n to 0?

“My first instinct was greedy, just keep subtracting the largest power of 2 that fits.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce software engineer interview, got hit with LFU cache as the main problem. Solid session technically but this one requires you to actually know the data structure cold, not just talk your way through it.
  • Design and implement an LFU cache with O(1) get and put operations, including proper eviction of the least frequently used key and LRU tiebreaking.

“I knew the high-level idea but the O(1) constraint is where it gets painful.”

View Post

Salesforce·Machine Learning Engineer·Onsite - Multi Round

Jul 2026
Salesforce ML Engineer loop, heavy on applied AI fundamentals and production systems. The questions ranged from textbook metrics to full LLM deployment architecture, so the depth required was pretty uneven and hard to pace.
  • What is F1 score and when should you prefer it over accuracy as an evaluation metric?
  • How do classification and regression differ in terms of objective, output type, and how you evaluate each?
  • How would you design a fair benchmarking process for an ML or LLM-based system?
  • What are common bottlenecks in model inference pipelines, and how do you optimize for latency and throughput?
  • Walk me through how you would take an AI agent service from prototype to production.
  • Explain the Transformer architecture at a high level, and then describe context engineering, retrieval-augmented generation, grounding, and guardrails as they apply to modern LLM applications.

“This felt like a warm-up but I over-explained it.”

View Post

Salesforce·Software Engineer·Technical Phone Screen

Jul 2026
Salesforce software engineer interview that went pretty deep into algorithm design. The core problem was grouping anagrams, which sounds straightforward until they push you to optimize past the obvious sorting approach.
  • Given an array of strings, group all words that are anagrams of each other. Start with a sorting-based solution, then optimize to linear time using character frequency hashing. Walk through key construction, collision handling, and the time/space tradeoffs between the two approaches.

“I started with the sort-each-word approach and got through it fine, O(n * m log m) and all that.”

View Post

Salesforce·Backend Engineer·Onsite - System Design / Architecture

Jul 2026
Salesforce backend round, one big system design question about elevators. Felt like a classic OOP design problem but with enough scheduling complexity to keep you on your toes for the full hour.
  • Design an object-oriented elevator control system for a building with multiple floors and multiple elevators. The system should handle hall requests from floors, cabin requests from inside elevators, track elevator state, assign requests to elevators, and schedule movement efficiently.

“Spent the first few minutes just listing classes out loud, ElevatorCar, FloorRequest, Dispatcher, Door, that kind of thing.”

View Post

Salesforce·Software Engineer·Onsite - System Design / Architecture

Jul 2026
Salesforce system design round focused entirely on building out a recommendation service end to end. Pretty dense session covering everything from API shape to storage schemas to how batch training and online serving share the same infrastructure without stepping on each other.
  • Design the overall serving infrastructure for a recommendation service, including how clients fetch data, the API design, and how you'd handle load balancing across multiple stateless application servers.
  • Walk through the storage schema design for users, items, and interaction logs, including how you'd structure fields like user_id, item_id, signal type, and timestamp.
  • How would you design indexing and partitioning strategies for the interaction logs to support both batch training jobs and low-latency online serving reads and writes?
  • How do you handle caching in this recommendation system, and what are the trade-offs between cache freshness and serving latency?
  • What approaches would you take to ensure fault tolerance and consistency in this system, especially given that batch training and online serving are running concurrently?

“Started with the API contract and worked outward, which felt right.”

View Post