A Prompter for the Technical Interviewer: Reksoft Built an AI to Catch AI Cheaters

The hiring market has turned into a battlefield: candidates use AI cheat sheets in technical screenings, with a neural network feeding the perfect answer into an earpiece during the interview itself. Reksoft responded asymmetrically by building an AI prompter for the interviewer. The tool analyzes the candidate's answers in real time and suggests non-standard follow-up questions that a canned answer has no response to.

AI-processed from Habr AI; edited by Hamidun News
A Prompter for the Technical Interviewer: Reksoft Built an AI to Catch AI Cheaters
Source: Habr AI. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceHabr AI ↗
Editor-in-Chief
Key metric #1Perfect syntax paired with an inability to explain the point in simple words
Key metric #2No concrete examples from personal experience
Key metric #3A pause at the worst possible moment, right before the "obvious" answer

Reksoft has published an article on Habr with an unexpected angle: while the market debates how candidates use AI in interviews, the company proposes arming the interviewer with the same technology.

The hiring market has changed radically

The flow of applications for IT jobs has grown many times over. The cause is neural networks, which in minutes generate résumés with CTO-level experience, cover letters and portfolios. HR departments are drowning: most applications are irrelevant, yet formally look flawless.

The problem gets worse at the technical screening stage. Tools marketed as "completely invisible" to the interviewer are being actively promoted. The scheme is simple: an AI listens to the interviewer's question in real time and instantly delivers a ready answer, to a second monitor, a smart headset or an earpiece. The candidate repeats it and comes across as a senior.

Why this doesn't work for long

An experienced technical interviewer spots the patterns of AI cheating faster than you might think. Syntactically flawless phrasing is the first signal. It is followed by characteristic pauses before a "ready-made" answer, a glazed stare at the monitor instead of natural eye contact, and complete bewilderment when faced with a non-standard follow-up question.

The author lists the markers he notices in real interviews:

  • Perfect syntax paired with an inability to explain the point in simple words
  • No concrete examples from personal experience
  • A pause at the worst possible moment, right before the "obvious" answer
  • Boilerplate phrasing reproduced exactly, time after time
  • Total disorientation at any deviation from the standard question
"You can tell when a robot answers your call to a support line and

when it's a live operator, right?" the author writes.

An AI cheat sheet can deliver a polished, formulaic answer. But it cannot react flexibly to the non-standard questions a live interviewer asks, and that is exactly its weak spot.

A prompter for the interviewer: an asymmetric response

This very principle, that a template has no answer to an unexpected question, underlies Reksoft's tool. The company developed an AI prompter for the technical interviewer. The tool runs in the background: it transcribes the candidate's answers in real time, analyzes their content and depth, and then suggests specific follow-up questions. The goal is to expose the emptiness behind a polished facade without turning the interview into a detective investigation.

According to its creators, the prompter does not replace the interviewer but strengthens them. The tool suggests question options, but the choice stays with the specialist. The logic is elegant: rather than trying to catch a cheater red-handed (technically difficult), create conditions in which the deception simply doesn't work.

What this means

The AI arms race has reached the hiring market, and this is only the beginning. Companies are adapting the format of technical interviews: strengthening practical tasks, introducing tools for interviewers, changing screening scenarios. Relying on a canned cheat sheet is getting harder, especially when a neural network is also working on the other side of the table.

Key Questions Answered

What are the key details on The hiring market has changed radically?

The flow of applications for IT jobs has grown many times over. The cause is neural networks, which in minutes generate résumés with CTO-level experience, cover letters and portfolios. HR departments are drowning: most applications are irrelevant, yet formally look flawless. The problem gets worse at the technical screening stage. Tools marketed as "completely invisible" to the interviewer are being actively promoted. The scheme is simple: an AI listens to the interviewer's question in real time and instantly delivers a ready answer, to a second monitor, a smart headset or

Why this doesn't work for long?

An experienced technical interviewer spots the patterns of AI cheating faster than you might think. Syntactically flawless phrasing is the first signal. It is followed by characteristic pauses before a "ready-made" answer, a glazed stare at the monitor instead of natural eye contact, and complete bewilderment when faced with a non-standard follow-up question. The author lists the markers he notices in real interviews: Perfect syntax paired with an inability to explain the point in simple words No concrete examples from personal experience A pause at the worst possible momen

What are the key details on A prompter for the interviewer: an asymmetric response?

This very principle, that a template has no answer to an unexpected question, underlies Reksoft's tool. The company developed an AI prompter for the technical interviewer. The tool runs in the background: it transcribes the candidate's answers in real time, analyzes their content and depth, and then suggests specific follow-up questions. The goal is to expose the emptiness behind a polished facade without turning the interview into a detective investigation. According to its creators, the prompter does not replace the interviewer but strengthens them. The tool suggests ques

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…