The Software Engineering Interview Has Changed: What interviewers are actually evaluating in 2026

The answer stopped being the hard part.

A correct solution now takes a model about eight seconds to produce. Google is piloting a coding round where you use Gemini and get scored on how you prompt it and whether you catch what it gets wrong. Amazon's recruiter guidance says generative AI may disqualify you. The policy changes from one round to the next inside the same company.

So what is left to evaluate?

Meera Kulkarni has five and a half years of backend experience, an order management system she has never been asked to explain, and three interview loops running at once. This book follows all three from the first recruiter call to the offer, and uses them to work out what interviewers are measuring now that the answer is free.

What is inside

• Researching a company so that it changes your answers, not your small talk
• Why retrieval practice feels worse than re-reading and works far better
• How much you need to know about AI when it is not your job, and what to skip
• What is actually scored when you use an assistant with someone watching
• Concurrency, failure, indexes and isolation levels, in any language
• How to ask an interviewer for what you need without sounding lost
• Stories specific enough that a sceptical interviewer could check them

Twenty thousand words. No problem catalogue, no study schedules, no scripts. Written for working backend engineers, with the bar marked wherever it moves for someone more senior.