Oracle interview experiences from 2025 new grad and IC3 loops

Oracle’s interview process for US new grads has two or three stages: a recruiter screen on your latest project, the technologies you know, graduation date and visa needs; one or two technical interviews of about an hour with LeetCode-style coding, sometimes on HackerRank (graphs, strings, hash maps, and in one loop a debugging session), mixed with questions about your resume; and on some teams a hiring manager round on your past work.

In the US, Oracle’s 2026 OCI Undergrad Software Engineer posting for Santa Clara lists a hiring range of $61,200-$141,200 a year, with possible bonus and equity. Experienced loops are longer: a July 2025 IC3 candidate had four rounds after the recruiter screen.

Oracle interview process, round by round

Oracle interview process for US new grads: 2025 reports

These are short candidate reports from Taro, not full write-ups, so read them as patterns.

• One hour, coding plus resume questions. A San Francisco new grad in February 2025 was asked by HR to fill in a basic information form, then had one 1-hour interview mixing coding with behavioral questions about resume projects. No offer.
• Recruiter screen on projects and logistics. A Software Engineer I candidate in April 2025 was asked about their most recent project and the technologies they knew, then about graduation date, visa requirements and the hiring teams. They did not move forward.
• HackerRank coding and a debug session. A US software engineer in September 2025 had a coding interview with common LeetCode-style questions on HackerRank and a debugging session, and got an offer.
• Two-part graph problem. A San Francisco software engineer in September 2025 was given a graph problem with a follow-up on how they would approach it, with strings and maps also covered, and accepted the offer.
• IC3, four rounds. An experienced software engineer in July 2025 had a hash map problem with behavioral questions, a manager round on current and past work, a priority queue problem, and 2D array coding with more behavioral questions. They declined the offer.

Oracle interview questions that come up often

Coding / technical (US reports)

1. A two-part graph problem, with a follow-up on how you would extend your approach (San Francisco, 2025)
2. A hash map problem, a priority queue problem and 2D array coding (IC3 loop, 2025)
3. Common LeetCode-style string and map problems on HackerRank
4. A debugging session: read existing code, find the bug and explain the fix
5. Your most recent project, the stack and the hardest decision in it

Behavioral and recruiter questions

1. Why Oracle, and which team: OCI, Database or Fusion Applications?
2. When do you graduate, and do you need visa sponsorship?
3. Tell me about a project you own and what you would change
4. A disagreement with a teammate and how you handled it

Oracle interview tips from 2025 reports

• Expect coding and resume questions in the same hour; the February 2025 new grad had one interview that mixed both.
• Practice reading and fixing unfamiliar code. One September 2025 offer loop had a debugging session next to the coding interview.
• Ask the recruiter which team the loop is for. SQL depth matters for Database and Fusion Applications roles more than for OCI.
• Have your graduation date, start date and sponsorship answer ready for the first call.

Hiring manager and behavioral questions

Oracle’s US loops put behavioral questions inside the coding interviews rather than in a separate HR round, and some teams add a manager conversation on your past work. Prepare a specific “Why Oracle?” that names the team, plus three project stories. Sample answers are on the Oracle behavioral interview questions page.

Sample answers for high-frequency questions

Q. Write a SQL query to find the departments whose average salary is above 50000.
A. Aggregate by department and filter on the aggregate with HAVING: SELECT d.dept_name, AVG(e.salary) AS avg_salary FROM Employee e JOIN Department d ON e.dept_id = d.dept_id GROUP BY d.dept_name HAVING AVG(e.salary) > 50000 ORDER BY avg_salary DESC. The key points an interviewer listens for are that every non-aggregated column in the SELECT must appear in the GROUP BY, that the JOIN happens before grouping, and that the filter on the average has to live in HAVING because WHERE is evaluated before rows are grouped. Logical execution order is FROM, then JOIN, then WHERE, then GROUP BY, then HAVING, then SELECT, then ORDER BY - knowing that order lets you answer most follow-ups.

Q. What is the difference between WHERE and HAVING?
A. WHERE filters individual rows before any grouping happens, so it cannot reference an aggregate function - writing WHERE AVG(salary) > 50000 is an error. HAVING filters the groups produced by GROUP BY, so it can reference aggregates like COUNT, SUM, or AVG. If a condition can be expressed in WHERE, put it there: it removes rows early and so reduces the work the grouping step has to do, which matters on large tables. A query can use both - for example WHERE join_date >= '2020-01-01' narrows the rows first, then HAVING COUNT(*) > 5 keeps only the departments that still have more than five such employees.

Q. How does an index speed up a query, and what does it cost?
A. Most relational indexes are B-tree structures that keep the indexed column’s values in sorted order with pointers to the underlying rows. Instead of scanning every row (a full table scan, O(n)), the database walks the tree in roughly O(log n) to find matching values, which is why an index on a column used in WHERE, JOIN, or ORDER BY can turn a multi-second query into a millisecond one. The costs are real: every INSERT, UPDATE, and DELETE must also maintain the index, and indexes consume disk space. Indexes also stop helping when the query applies a function to the column (WHERE UPPER(name) = 'X' cannot use a plain index on name), when the column has very low cardinality, or when the query returns a large fraction of the table anyway.

Q. What is the difference between a primary key, a unique key, and a foreign key?
A. A primary key uniquely identifies each row, cannot contain NULLs, and a table can have only one - most databases create a clustered or unique index on it automatically. A unique key also enforces uniqueness but permits NULLs (how many NULLs is allowed varies by database) and a table can have several, which is how you enforce that email and phone are each unique alongside a numeric primary key. A foreign key is a column in one table that references the primary or unique key of another, enforcing referential integrity so you cannot insert an order for a customer_id that does not exist. Foreign keys also drive ON DELETE CASCADE or ON DELETE SET NULL behavior when the parent row is removed.

Q. When would you deliberately denormalize a database?
A. Normalization removes redundancy and update anomalies, but every extra table means another join at read time. You denormalize when read performance matters more than storage and write simplicity - a reporting or analytics table that would otherwise join six tables, a product listing that stores a cached review_count instead of recomputing COUNT(*) on every page load, or a data warehouse star schema where dimensions are intentionally flattened. The trade-off is that duplicated data can drift out of sync, so you need a disciplined way to keep it correct: triggers, application-level updates, or a scheduled refresh. The interview-safe answer is: normalize by default to 3NF, then denormalize selectively with measurements showing the join is the actual bottleneck.

Oracle coding, interview and recruiter-call tips

Oracle India: campus interview experiences, eligibility and pay

Oracle’s India campus process includes: 1. Online Coding Round (90-120 minutes) - 2-3 coding questions on DSA and database concepts. 2. Technical Interview Round 1 (45-60 minutes) - DSA and algorithms. 3. Technical Interview Round 2 (45-60 minutes) - System design and database concepts. 4. HR Interview (30 minutes) - Motivation and company fit. Total duration: 2-3 weeks from application to offer.

Round-wise process (India campus)

Real interview experiences from India

These are student write-ups for Oracle’s India campus roles (Application Engineer / ASE). Candidates who prepare only LeetCode are often surprised by how much of the loop is MCQ aptitude, puzzles, C-level memory questions and DBMS rather than contest-style algorithm problems.

Experience 1: Application Engineer, on-campus (selected)

Source: Oracle Interview Experience for Application Engineer (On-Campus) - GeeksforGeeks

Round 1 - online test (2 hours, MCQ)

• Not a coding round. Sections covered numerical aptitude, comprehension, reasoning, data structures (mostly AVL trees and code-snippet output questions), and software engineering.

Round 2 - technical interview 1

• Introduction, then puzzles including the three-ants-on-a-triangle problem, a star-pyramid pattern, and arranging a list of numbers to form the smallest number without sorting.
• OS (threads), a DBMS question, and arrays vs linked lists.

Round 3 - technical interview 2

• Types of joins, ACID properties, three more puzzles explained aloud, OOP with examples, interfaces vs abstract classes.
• Linked list initialization with pointers, then memory: new vs malloc, free, destructors, memory leaks, dangling pointers, and code to allocate memory for four integers.

Round 4 - technical + HR

• Tree types and a Red-Black tree explained and coded, machine learning basics, Oracle’s autonomous database, REST APIs with code to call one, threads, Linux commands, cloud concepts, static variables, and a screen-shared walk through one of the candidate’s own GitHub repositories.
• A coding task: read usernames from a file, call an API that lists event attendees, and print the matches.
• HR: why Oracle, family, other offers and the reasons for them, hackathons, and an ER diagram for a hackathon project’s database.

Result: Selected.

What other Oracle campus accounts add

• The process shape: an online test, two technical interviews and a combined technical + HR round in the Application Engineer and ASE write-ups; a 2024 account had three technical rounds instead, the last led by a VP.
• Online tests can include an API question alongside DSA and aptitude (ASE 2023 and the 2024 account).
• SQL and DBMS come up in most rounds: SQL queries (ASE), joins and normalization, and when to denormalize (2024).
• Puzzles still appear: a magic square with missing values, Camel and Banana, and the Monty Hall problem. Think aloud rather than going quiet.

Sources: Oracle campus placement - Application Engineer · Oracle Interview Experience for ASE (On-Campus) · Oracle Interview Experience (On-Campus) 2024

Takeaways

• What decided it: Defending SQL/DBMS aloud (joins, normalization, ACID) alongside DSA, and thinking through puzzles without going quiet
• Common fail: Treating SQL as optional or assuming managerial is soft HR
• How to practice: Join/normalization drills; keep DSA warm through every round

Oracle India campus questions

1. Tree variants: AVL tree and code-snippet output questions in the test, Red-Black trees with code in the interview
2. SQL queries, joins, normalization and ACID
3. C memory management: malloc/free, memory leaks and dangling pointers
4. Puzzles such as three ants on a triangle, a magic square, Camel and Banana, and Monty Hall
5. Matrix coding problems: find a word in a character grid, traverse a matrix in a staircase pattern

Past campus papers are on Oracle on placementpapers.app.

Sources

US reports come from Taro: a San Francisco new grad, a Software Engineer I screen, HackerRank coding and debugging, a San Francisco graph problem and an IC3 loop. US pay is from Oracle’s 2026 OCI Undergrad Software Engineer posting.

Similar companies

SAP · Microsoft · IBM · Zoho · Freshworks · Salesforce