Company-Specific Guide

Citi Karat Interview: Format, Questions & Prep (2026)

LeapFork is an independent interview preparation platform, not affiliated with Citi or Karat, Inc. This guide is based on candidate-reported experiences and publicly available information.

Citigroup (Citi) — one of the world's largest financial institutions — uses Karat for technical screening of software engineering and technology candidates in its global technology division. Citi's technology teams build systems for trading, risk management, digital banking, payments, and data analytics at institutional scale.

LeapFork is an independent prep platform, not affiliated with Citi or Karat. This guide is based on candidate-reported experiences. Citi's engineering culture operates in a regulated, high-reliability environment — correctness and thoroughness matter even more than speed. This shows up in how candidates should approach edge cases: think about data integrity and error conditions explicitly.

How Citi Uses Karat

The Citi Karat screen follows the standard 60-minute format: introduction, domain questions, and approximately 40 minutes of live coding.

Domain questions for Citi technology roles may touch on software engineering fundamentals with a slight lean toward reliability and correctness — how you handle errors, how you think about data validation, what you would do if an API returned unexpected data. These questions reflect the financial services context where an incorrect calculation can have significant consequences.

The coding segment follows Karat's standard progressive structure. The problem difficulty is consistent with other Karat client companies (easy-to-medium). Citi does not appear to use finance-specific mathematical problems in the Karat format.

Role Levels & What to Expect by Level

Citi uses Karat for software engineering roles across its technology division, including roles in application development, data engineering, and platform teams. The levels screened via Karat vary — the Karat format appears most commonly for analyst through vice president level technical roles (Citi uses a banking-style leveling structure).

Technology roles at Citi span a range of seniority and specialization. Your recruiter will clarify whether your specific role uses the Karat format or a different assessment approach. Many Citi technology roles that candidates apply to through platforms like LinkedIn and Citi's careers portal use Karat as the first technical screen.

Question Patterns to Practice

Based on candidate reports, Citi Karat problems follow the standard Karat distribution. Finance-specific mathematical knowledge is not tested in the coding segment.

Patterns to prioritize:

Hash map state tracking: The core Karat pattern. Practice frequency maps, two-sum variants, and grouping-by-key problems until the approach is automatic.
String and data parsing: Financial systems handle a lot of structured text (transaction descriptions, account identifiers, formatted amounts). String parsing problems are a natural fit.
Array manipulation: Sorted arrays, prefix sums, simple aggregations.
Error handling thinking: While not a coding pattern, be prepared to verbally discuss how your code handles invalid input, null values, or boundary conditions. This matters more in a finance context.

When answering domain questions, frame your thinking around reliability and correctness — not just efficiency. In financial software, a correct O(n²) solution that handles all edge cases is often better than a theoretically faster solution that misses corner cases.

Practice the Real Thing — Before the Real Thing

LeapFork's timed simulation mirrors the actual Karat format: live coding, AI-graded performance, and a speaking AI interviewer. Free to start.

Run a Free Simulation →

2-Week Prep Plan

Week 1: Core patterns with correctness focus

Days 1–2: Hash map problems. Focus on getting the edge cases right (empty input, duplicates, single element) — not just the happy path.

Days 3–4: String parsing. Practice handling malformed input gracefully. Ask "what if the input is empty or has unexpected characters?" in every problem.

Days 5–7: Mixed array + hash problems. After solving each, explicitly state: "What would break this code? How would I handle it?"

Week 2: Pacing and communication

Days 8–10: Timed 40-minute sessions. After each: how many parts did you complete? Did you narrate edge case handling out loud?

Days 11–12: Domain question practice. Prepare concise answers about error handling, data validation, and testing. Practice these verbally.

Days 13–14: Two full timed simulations. Camera on. Debrief edge case coverage after each.

Frequently Asked Questions

Does the Citi Karat screen include finance or math questions?

Based on candidate reports, the Citi Karat coding screen uses standard algorithmic problems, not finance-specific math or domain knowledge. Domain questions may touch on general software engineering concepts with a reliability focus. You are not expected to know financial instruments or accounting.

What programming language should I use for the Citi Karat screen?

Citi's technology division uses Java, Python, C++, and other languages depending on the team. For the Karat screen, use your strongest language. Java is a common choice for candidates with enterprise backgrounds; Python is concise for algorithmic problems.

What happens after the Karat screen at Citi?

Candidates who pass the Karat screen at Citi typically proceed to interviews with Citi technology leaders and engineers. The process varies by role, team, and business line. Your recruiter will outline the subsequent steps for your specific position.

Last reviewed: 2026-07-01. “Citi” is a trademark of its respective owner, used here descriptively to identify the interview process.