Freshworks Interview Questions and Answers
Last updated 2026-11-19
Freshworks builds SaaS products such as customer support, CRM and IT service tools, so its engineering questions favour multi-tenant design, ticket and workflow modelling, API quality and operational thinking, next to medium-level data structures problems. This page offers practice questions with worked answers across those areas and the behavioural round.
They are practice questions on topics the rounds are commonly reported to cover, not a record of any real interview. Code problems are in Freshworks coding round questions. The process is outlined in how to crack the Freshworks interview.
| Typical rounds | Coding/DSA, system design (senior), leadership (senior) (commonly reported) |
|---|---|
| Roles | Software Engineer |
Rounds as commonly reported
Candidate reports describe a coding or DSA round with medium LeetCode-style problems plus resume discussion, system design for senior engineers, and a leadership or behavioural round for senior roles.
Freshworks does not publish its rounds, so these vary by team and level.
Hiring status
An Outlook Business report says Freshworks cut about 11 percent of its workforce, around 500 people across India and the US, which it describes as the company's third major layoff round, after an earlier cut of 13 percent (about 660) in November 2024. Read Freshworks hiring for more, and use the company's careers page at freshworks.com/company/careers for current openings.
How to prepare
Choose one of the Freshworks products, such as a helpdesk, and write down its nouns (ticket, agent, customer, SLA, automation rule). Then practise designing its data model and the main flows; it is the fastest way to find the questions behind the questions.
Freshworks Interview Questions and Answers: questions with worked answers
1 What is multi-tenancy and how would you implement it for a SaaS helpdesk?
mediumMultiple customers (tenants) share the same application while their data stays isolated. Options: a shared database with a tenant id column on every table and enforced filters (cheapest, needs strict discipline), a schema per tenant, or a database per tenant (strongest isolation, higher cost). Choose by tenant size and compliance needs, always scope every query by tenant, and watch for noisy neighbours with per-tenant rate limits and quotas.
2 Design a ticketing system data model.
mediumTables: Ticket (id, tenant, requester, assignee, status, priority, created, due), Comment, Attachment, Tag, and an Event log of status changes. Index tenant plus status plus updated time for agent queues and tenant plus requester for customer history. Keep a state machine for status (New, Open, Pending, Resolved, Closed) so transitions are validated, and store events immutably to compute SLA reports.
3 How would you implement SLA timers that pause when a ticket is pending on the customer?
hardStore the time spent in running states, not just created time. On each transition, close the current running interval and open a new one if the new state counts toward SLA. Compute the due time from remaining allowance and business hours calendar. A scheduler checks upcoming breaches from an indexed due-time column and emits alerts before and at breach.
4 Design an automation rule engine ("if priority is urgent and no reply in 1 hour, notify manager").
hardRepresent a rule as conditions plus actions in a structured format. On each ticket event, a worker loads the tenant's rules, evaluates conditions, and runs actions through an action interface (assign, tag, send email). Time-based conditions use scheduled checks keyed by ticket. Prevent loops by limiting rule chains per event, and log each rule run for debugging.
5 How do you design a REST API that is easy to evolve?
mediumVersion in the path or header, make additive changes only within a version, use consistent resource naming and HTTP verbs, paginate with cursors, return clear error codes and messages, and support idempotency keys for creating resources. Document with an OpenAPI schema, add rate limiting with headers telling the client its quota, and deprecate with a timeline and warnings.
6 Explain the difference between SQL and NoSQL for storing ticket conversations.
mediumTicket metadata fits relational tables because you query it by many fields and join to users and teams. Conversation threads are append-heavy and read by ticket id, suiting a document or wide-column store, or just a table indexed by ticket id. Full-text search should go to a search engine. Many products use all three with a clear source of truth.
7 How would you add full-text search over tickets with tenant isolation?
hardIndex tickets in a search engine with the tenant id as a mandatory filter (or separate indices for the biggest tenants). Update the index from change events, accept brief lag, and verify permissions again when returning results. Support filters for status and assignee through facets, and rebuild the index from the database when the schema changes.
8 What are background jobs, and how do you make them reliable?
mediumWork outside the request path such as sending emails or importing data. Put jobs on a durable queue, make handlers idempotent because jobs can run twice, set retries with backoff and a maximum, route failures to a dead-letter queue, and add metrics on queue depth and age. Long jobs should checkpoint progress.
9 Explain optimistic concurrency for two agents editing a ticket at once.
mediumEach ticket carries a version number. An update says "change field X, assuming version N". The database applies it only if the version still equals N and increments it, otherwise the client gets a conflict and reloads or merges. This avoids locks and lost updates. For fields like comments, append-only data avoids conflicts entirely.
10 How do you monitor a SaaS service in production?
mediumTrack the golden signals: latency, traffic, errors and saturation, per endpoint and per tenant tier. Add structured logs with a request id, distributed traces, and alerts based on symptoms users feel (error rate, slow responses) rather than only on CPU. Keep runbooks, and review incidents blamelessly to fix causes.
11 Tell me about a time you led without formal authority.
easyUse STAR. Describe the goal, how you built agreement (data, a prototype, listening to objections), how you divided work, and the result. Mention one thing that did not go smoothly and what you changed.
12 How do you handle changing priorities, especially in a company that is restructuring?
easyStay calm and practical. Ask what has changed and why, clarify what the new top priority is, close or hand off in-flight work cleanly, and communicate impact on timelines early. Show that you can work with uncertainty without losing quality.
Frequently asked questions
What does Freshworks ask in system design?
Reports mention system design for senior roles. SaaS themes such as multi-tenancy, queues and search are natural fits.
How hard is the Freshworks coding round?
Candidate reports say medium LeetCode-style problems plus discussion of your resume.
Is there a leadership round?
Reported for senior roles. Prepare stories on influence, conflict and ownership.
How do I find Freshworks openings?
On the company's careers page, and see Freshworks hiring for sourced news.
Sources
Related guides
More in Company Interview Questions
- Goldman Sachs coding round questions with solutions
- Goldman Sachs interview questions with answers (engineering analyst)
- Google Coding Round Questions with Solutions in Python and Java
- Google Interview Questions and Answers for SDE Roles
- Groww Coding Round Questions With Solutions
- Groww Interview Questions and Answers
- HCLTech Coding Round Questions with Solutions
- HCLTech Interview Questions for Freshers
- Hexaware Coding Round Questions with Solutions
- Hexaware Interview Questions for Freshers
- Honeywell Coding Round Questions with Solutions
- Honeywell Interview Questions and Answers for India Engineering Roles