How to nail a technical interview, according to a founder who's interviewed software engineers for Pinterest, Intuit, and Compass

How to nail a technical interview, according to a founder who's interviewed software engineers for Pinterest, Intuit, and Compass

Jeffrey Spector

Courtesy of Jeffrey Spector

Jeffrey Spector, cofounder of Karat.

  • Jeffrey Spector is the cofounder of Karat, a startup that conducts first-round technical interviews on behalf of tech companies such as Pinterest, Indeed, Intuit, and Compass.
  • The company's first-round interviews incorporate a problem-solving assessment, which is common within the interview process for engineers.
  • Spector's advice for young engineers looking to nail the interview is to come prepared with questions to ask the interviewer as well as talk through the process when trying to solve a problem.
  • He also suggested that engineers apply again if they're rejected from a major tech company.
  • Click here for more BI Prime stories.

In 2014, Jeffrey Spector wanted to help people get jobs more easily.

At the time, Spector was working as the chief of staff to Melinda Gates at the Bill & Melinda Gates Foundation. He was studying hiring across industries and observed that most interview processes were inflexible and inconsistent for candidates. So he thought about what an experience would look like if candidates were treated as well as customers. He partnered with Mohit Bhende, a Microsoft executive who had also observed that engineering teams were taking as long as 100 hours to make a single hire. The pair felt that interviewing needed to be treated as a full-time profession.

Today, Spector and Bhende are the cofounders of Karat, a startup that conducts first-round technical interviews on behalf of leading organizations such as Pinterest, Indeed, Intuit, and Compass. Frequently, these companies find that they are forced to sacrifice product development to interview and hire software engineers, or vice versa - so they approach Karat to save time and money. To date, the company has raised $41.6 million in funding, and its last funding round, led by Tiger Global Management, brought in $28 million.

How Karat works

Karat strives to provide companies with consistent candidate evaluations.


"For teams, the consistency means confidence that engineers will meet their hiring bar and that they don't miss out on great candidates who would have otherwise been overlooked," Spector said. "It's absolutely critical that you have a consistent first-round technical interview because it's a big variable that is then controlled. This helps identify the drivers of candidate performance in the onsite interview, so you can really ensure your team is interviewing for relevant skills and can accurately predict who will do well onsite."

The company's first-round interviews incorporate a technical, problem-solving assessment, common within the interview process for engineers. What's not as common, however, is that Karat's interviewers are software developers who have made interviewing their job - they call them "Interview Engineers"- and many come from major tech companies like Amazon and Facebook.

"We realize interviewing is stressful, and when you're stressed it's hard to put your best foot forward. Interview Engineers are equipped with best practices for how and when to most effectively provide … guidance and clarity in the technical interview," said Spector. Plus, Karat allows candidates to interview around the clock at a time that best suits them, and even the option to redo the interview if they feel they could have done better. "Just knowing this is an option reduces anxiety for the candidate," he added.

After Karat conducts the first-round technical interview, they deliver a recommendation to the company's hiring team, who then makes the final hiring decision. The startup also collects data from the entire candidate lifecycle to evaluate and learn from hiring outcomes.

After reviewing tens of thousands of interview loops and observing the success of both candidates and the interviewing teams of hiring companies, Spector has identified clear insights for software engineering job candidates looking to get a leg up in their next interview at a major tech company.


Do your research and come prepared with questions

Spector emphasized that when applying for a tech job, it's key to keep an open mind. "Don't be too overly focused on the brand-name companies" like Facebook, Amazon, Apple, Netflix, and Google, he said. "We see a lot of people starry eyed … but there are a lot of amazing tech companies out there," Spector said.

He added that candidates should understand their level of technical ability and apply to roles at companies where their skills are a match. "Some companies do require demonstrable skills that take time to learn, like mobile. This will usually be identified in a job description as a basic qualification," Spector explained.

It's also important to think about and research the company's core business and go into the interview with plenty of knowledge about their product and mission. Spector has seen feedback notes from candidates' onsite interviews that stated they were knocked out because they hadn't thought about the product or the scale on which the company was operating. He also has seen notes highlighting that some candidates didn't have any questions to ask at the end of an interview.

"Understand who you're interviewing with and what problems they're solving on a daily basis. Ask questions about their job," Spector advised. "For example, if you are interviewing with an engineering leader, you might ask them about hiring goals, prioritization methodologies, and the team's skill gaps. Most candidates won't ask these questions, but they demonstrate an interest in the manager's responsibilities and show that you may have management potential."

Talk through your thought process when solving a coding problem

In a typical engineering interview, you may be asked to solve coding problems on a whiteboard, piece of paper, or computer, without the assistance of tools you'd normally have. "Practice problems with a timer and understand how to solve problems effectively," Spector suggested. He also advised that candidates brush up on core algorithms and the languages they are most comfortable using to code.


"We see a lot of candidates fall down when their chosen language in the interview is not one they're very familiar with," he said. Spector recommends that candidates know how to interpret error messages and warnings in their language of choice. "No matter how much you prepare, you will probably make some mistakes, but knowing how to easily recover from those mistakes is a key factor in your performance," he added.

While solving a problem in any interview, candidates "should communicate openly about what their thought process is and how they go about problem solving," Spector said. He has seen some candidates begin to solve problems during the onsite interview without clarifying what they've been asked, which wastes valuable time.

"Even though you may feel vulnerable, don't be afraid to ask questions. They can actually help you," said Spector. "Interviewers at tech companies often deliberately hold back information to see if you can effectively scope problems." (However, Spector noted that this is not Karat's practice: "Interview Engineers at Karat are very transparent with candidates about what they are expected to demonstrate when solving a problem in the technical interview.")

Communicating your thought process offers you two distinct advantages. The first is that if you're solving a problem incorrectly, you might get some help from the interviewer to course correct. The second is that even if you run out of time and don't finish your implementation, or only partially solve the problem, the interviewer may be able to give you credit for the direction you're heading. This is likely, as Spector pointed out that "100% completeness isn't usually the goal. What's more important is that you articulate the remaining steps in your solution and the difficulties you would anticipate."

On the flip side, if you don't communicate, the interviewer may not be able to tell whether the approach you were going to attempt would have been valid. You "can get the majority of credit for that problem by verbalizing [your] understanding," Spector said.


If at first you don't succeed, try again

The best tech companies, Spector said, are also "building communities of talent [and] innovating on how to build long-term relationships in a [tight labor] market."

So, he added, "even if you get rejected today, you could be hired in six months or one year." He pointed to companies like Citrix that hire what they call "silver medalists" - even filling 15% of their non-referral roles with candidates who reapplied.

If candidates don't land their ideal position on the first try, they should work on building their skill set and keep in touch with the companies they're interested in working for.

Spector's final piece of encouragement for candidates is that "every interview, regardless of the outcome, is a learning opportunity. If you are asking for feedback, listening … and reflecting on your performance, each interview should bring you closer to your ideal job."

Hannah H. Kim is an independent business and tech journalist. She is from Los Angeles and a graduate of the Iowa Writers' Workshop.