A technical interview tests whether you can do the technical work of the job and how you reason while you do it. Hiring processes for software, data and engineering roles usually combine several formats: a technical screen, one or more live coding or problem-solving sessions, a system design discussion for experienced candidates, and sometimes a take-home assignment.
To prepare, find out which formats you will face, review the fundamentals they draw on, practice timed problems out loud, and rehearse explaining the projects on your resume.
Key takeaways
- Ask the recruiter which formats and topics to expect. It is a normal question, and the answer decides what you study.
- Interviewers score your reasoning and communication, not only the final answer.
- Practice aloud and against a clock. Solving silently at your own pace does not train what is tested.
- Use the same routine for every problem: restate it, try an example, agree on a plan, code, test.
- Know every project on your resume well enough to defend the decisions in it.
- Rules on AI tools differ between employers. Check before any assessment or take-home.
01
What are the formats of a technical interview?
| Format | What happens | What it tests |
|---|---|---|
| Technical screen | A short call with an engineer or recruiter, often with concept questions | Fundamentals and whether a longer interview is worthwhile |
| Online assessment | Timed problems on a testing platform, often scored automatically | Correct, efficient solutions under time pressure |
| Live coding | You solve a problem in a shared editor or on a whiteboard while an interviewer watches | Problem solving, code quality and how you communicate |
| System design | An open-ended prompt to design a service or feature | Requirements, trade-offs and judgment at scale |
| Take-home assignment | A small project completed on your own time | How you write and organize real code |
| Project deep dive | Detailed questions about work on your resume | Whether you understand what you built and why |
Virginia Tech’s career office notes that many online assessments are scored on correctness, efficiency and completion within the time limit, and that most run 45 to 90 minutes. For data roles, expect the same formats with different content: MIT’s career office describes data science interviews as coding in Python or SQL, discussing when to use which model, and interpreting data or working through a case.
Role-specific question lists are in the guides to software engineer interview questions and data analyst interview questions.
02
What do technical interviewers evaluate?
Not just whether you reach the right answer. Virginia Tech’s career office lists what interviewers want to see: how you approach the problem, how you communicate your thinking, how you apply what you know, and how you respond when you get stuck.
In practice an interviewer is filling in answers to questions like these.
- Did the candidate understand the problem before starting, and ask about inputs and edge cases?
- Did they consider more than one approach and explain the trade-off?
- Is the code correct, readable and tested?
- Could I follow their reasoning the whole way through?
- When I gave a hint, did they use it?
That last point is easy to overlook. As MIT’s advisers note, interviewers can offer hints if you let them know your thought process, and sound reasoning can still leave a good impression when the final answer falls short.
03
A four-week study plan for technical interviews
Adjust the length to your starting point. Someone who solves problems daily at work may need a week of review; someone returning after years away may need more than four.
| Week | Focus | What to do |
|---|---|---|
| 1 | Map the interview and rebuild fundamentals | Ask the recruiter about formats. Choose one language. Review arrays, strings, hash maps, sets and how to reason about time and space cost. Solve one or two easy problems a day, aloud. |
| 2 | Core patterns | Trees, graphs, recursion, sorting and searching. Learn to recognize which approach a problem calls for instead of memorizing solutions. Start timing yourself. |
| 3 | Your format and your resume | System design, SQL, statistics or whatever your role adds. Prepare a five-minute walk-through of two or three projects. Do one mock interview. |
| 4 | Full simulations | Two or three complete practice interviews with another person, timed, with feedback. Revisit weak areas. Rest the day before. |
A daily session that works: one new problem under a time limit, one old problem solved again from scratch, and ten minutes explaining a concept aloud as if to an interviewer.
If you have only a few days, cut the plan down in this order: confirm the format, practice the routine below on five or six representative problems, and rehearse your project walk-through.
04
How to think aloud in a coding interview
Use one routine for every problem so that it holds up when you are nervous.
- Restate the problem in your own words and ask clarifying questions: input size, empty or repeated values, what to return when there is no answer.
- Work a small example by hand. It confirms that you understood, and it often shows you the method.
- Describe an approach before you write anything. Start with the simple one and its cost, then say how you would improve it. Get the interviewer’s agreement.
- Write the code in small steps. Narrate intent, not syntax: “this loop finds the first position where …”
- Test it. Trace a normal case and an edge case through the code line by line, out loud.
- Sum up. State the time and space cost and what you would change with more time.
It is fine to ask for a moment to think before you start. Silence for half a minute with a warning is better than talking without direction.
Starting: "Let me check I understand. The input is [...] and I need to return [...]. Can the input be empty? Can values repeat?" Planning: "The simple way is [...], which costs [...]. I think I can do better with [...]. Should I start with the simple version or go straight to the faster one?" Pausing: "I'd like thirty seconds to think this through before I write anything." Stuck: "I'm stuck on [specific part]. What I know so far is [...]. I see two options, [A] and [B]. Let me try [A] on a small example." Using a hint: "That helps. If I [apply the hint], then [consequence], so I'll change [...]." Finishing: "Let me test this with [case]. With more time, I'd improve [...]."
05
How to approach a system design interview
System design questions are open on purpose. The interviewer wants to see you turn a vague request into a reasonable plan, so lead the conversation.
- Clarify the requirements. Who uses it, what must it do, and what can be left out?
- Estimate the scale roughly: users, requests, data volume. Say your assumptions aloud.
- Define the interface and the data. The main operations and what is stored.
- Sketch the main components and walk one request through them.
- Find the weak points. Where does it slow down or fail first, and what would you do about it?
- Explain the trade-offs behind each choice, and what you would revisit as the system grows.
There is rarely a single right design. Stating your assumptions and comparing options shows more than reciting an architecture from memory.
06
How to handle a take-home assignment
- Read the instructions twice and ask questions early. Confirm the deadline, the expected time and how it will be judged.
- Respect the time limit. Finish the core well, and list what you would do next instead of doing it badly.
- Make it easy to run. Include setup steps, tests for the main logic and a short note on your decisions and trade-offs.
- Expect a follow-up. Employers often ask you to walk through the code and extend it live, so submit only what you can explain.
- Check the rules on AI tools. Virginia Tech’s career office advises asking the recruiter, because expectations vary widely between companies. If tools are allowed, test what they produce and be ready to explain every line.
07
How to prepare for the non-coding questions
Technical interviews usually include a conversation about your past work and some behavioral questions. For each project on your technical resume, prepare the problem, your approach, the tools, the result and what you would do differently. Have a few stories ready for teamwork, disagreement and failure; behavioral interview questions lists the common ones. For remote sessions, test the coding environment and your setup as you would for any video interview.
FAQ
Frequently asked questions
How long does it take to prepare for a technical interview?
It depends on how recently you have done this kind of problem solving. Several weeks of steady daily practice is a reasonable plan if your fundamentals are sound; allow longer if they are rusty. Take one timed practice problem on day one to see where you stand, and plan from there.
Which programming language should I use in a coding interview?
Use the language you know best, as long as the employer allows it. Fluency matters more than fashion, because you need your attention for the problem and not the syntax. If the job posting centers on a specific language, ask the recruiter whether you are expected to interview in it.
What should I do if I cannot solve the problem?
Keep reasoning aloud. State what you understand, solve a simpler version or write the straightforward solution first, and say where you are stuck. Interviewers can give hints when they can follow your thinking, and clear, logical progress on a hard problem is easier to credit than silence followed by a guess.
Can I use AI tools in a technical interview?
Only if the employer says you can. Policies differ from company to company and sometimes between stages, so ask the recruiter before an online assessment, a take-home or a live session. Where tools are permitted, you are still expected to verify the output and explain it.
How is an online coding assessment scored?
Many are scored automatically. Virginia Tech’s career office describes the usual basis as correctness, efficiency and completion within the time limit. Practice under timed conditions on a similar platform beforehand, read each question fully before coding, and test against the edge cases the problem statement mentions.
Sources
- Virginia Tech Career and Professional Development: Technical Interviewing: interview formats, the routine for coding problems, topics to review, how online assessments are scored and checking the rules on AI tools.
- Virginia Tech Career and Professional Development: Interview Types: what technical interviewers look for beyond the right answer.
- MIT Career Advising & Professional Development: Ace the data science technical interview: what data science interviews cover, taking a moment before answering, and voicing your thought process so interviewers can give hints.
Published by Jobbie and last updated on October 6, 2026. This guide is general information for job seekers, not legal, tax or financial advice. Spotted something wrong? Tell us.