I’m seeing a growing issue in tech hiring that’s honestly getting out of hand — AI-assisted / proxy interviews.
Candidates are using:
- Real-time AI tools to generate answers
- Hidden setups where someone else feeds responses
- Memorized + AI-polished system design answers
And the result?
👉 People with little to no real production experience are clearing interviews
👉 Meanwhile, genuinely experienced engineers are getting filtered out
-–
### What’s frustrating
As someone with real hands-on experience (Kafka, Spark, pipelines, etc.), interviews are becoming less about:
- Debugging ability
- Real-world problem solving
- Tradeoff thinking
…and more about:
- How fast you answer
- How polished you sound
- How “perfect” your responses are
Which is exactly what AI tools optimize for.
-–
### The real problem
Most interviews are still:
- Predictable
- Theory-heavy
- Easily searchable
So AI can:
- Suggest structured answers instantly
- Fill gaps in knowledge
- Make candidates sound more experienced than they are
But this breaks the hiring signal completely.
-–
### What companies don’t realize (yet)
You can pass interviews with AI help…
But you can’t survive real production issues with fake experience.
Things like:
- Kafka data loss / ISR issues
- Spark skew / memory tuning
- Airflow pipeline failures
- Debugging distributed systems
These require actual experience, not generated answers.
-–
### Impact on genuine candidates
- You might hesitate → AI-assisted candidates don’t
- You think → they respond instantly
- You speak from experience → they speak in “perfect frameworks”
So ironically, real engineers look weaker in interviews
-–
### What needs to change
If companies want to fix this, interviews should shift toward:
✔ Deep dives into actual past work
✔ Failure scenarios (“what broke and how did you fix it?”)
✔ Follow-up drilling (keep going until depth is exposed)
✔ Live problem solving (not memorized questions)
Because AI can fake answers…
…but it can’t fake experience under pressure
-–
### Curious to hear from others
- Are you seeing this in your interviews?
- Have you encountered obvious proxy/AI-assisted candidates?
- How are companies adapting (if at all)?
This feels like a serious issue that’s only going to grow if ignored.
