Interview Solver is an AI interview co-pilot desktop application for software engineers. It provides real-time LeetCode problem-solving assistance, system design answers, stealth anti-detection mode, screenshot problem-solving, voice transcription and other functions to help job seekers instantly obtain problem-solving ideas and code solutions during technical interviews. Covers the interview question bank of first-line technology companies such as Google, Amazon, Meta, Microsoft, etc.

InterviewSolver

Core parameters and statistics of Interview Solver

Interview Solver is an AI Interview Copilot desktop application for software engineers. It is officially positioned as "Your Unfair Advantage in Coding Interviews" - it provides real-time problem-solving assistance during technical interviews and is invisible to screen sharing. The product belongs to Type D (productivity/business application), and the core delivery form is a native desktop client (Windows/macOS), not a browser extension or web application.

Projects	Public Information
Product positioning	AI interview co-pilot desktop application, real-time programming interview assistance
Core Competencies	LeetCode full question bank problem solving, system design answers, screenshot problem solving, voice transcription, stealth anti-detection
Product form	Desktop client (Desktop)
Business Model	Free Trial + Monthly Subscription ($39/mo)
Target users	Software engineers, technical interview job seekers
User scale	Official disclosure "Trusted by 10428+ engineers"
Recommendation rate	Official disclosure "98% Would recommend"
Home	US
Supported platforms	Windows, macOS (desktop, not web)
Supported languages	en-US
Competing products	Final Round AI ($149/mo), UltraCode, Interview Coder, LockedIn AI, AI Apply
Free Tools	ATS Resume Checker STAR Method Builder, Salary Query Big O Cheat Sheet, Total Contract Calculator

Interpretation of key parameters: The differentiation of Interview Solver lies in the "native desktop client" rather than the browser extension. This means it can seamlessly switch between the foreground coding window and the background AI window via global hotkeys without the need for Alt+Tab to switch tabs - the latter of which can be easily noticed by interviewers in screen sharing scenarios. The official claims that it is completely invisible to Zoom ≤6.16 and is hidden in Activity Monitor. This is its core barrier compared to competing Chrome extension products.

User Scale: The official page discloses "10428+ engineers trust us". Judging from the early stages of tool-based products, this user base means that the product has had certain market verification, but has not yet entered the stage of large-scale popularization.

Interview Solver’s users and market recognition

Interview Solver is in the AI interview assistance track, a segment that will grow rapidly in 2025-2026. The core driving force is the intensifying competition in technical interviews among technology companies and the continued heating up of LeetCode’s question-brushing culture.

Market Positioning: Directly competes with Final Round AI ($149/month). Interview Solver uses $39/month pricing and the invisibility capabilities of the native desktop client as its core points of differentiation. On its official comparison page, it claims to be ahead of Final Round AI in 16 key functional dimensions. In addition, it also forms a competitive matrix with tools such as UltraCode, Interview Coder, LockedIn AI, and AI Apply.

User testimonials: Officially displayed multiple user testimonials, including "Cracked the SDE 2 interview at Microsoft", "Used it to pass my phone screen", "Helped me land my dream job at a FAANG company", etc. The official page has an average public rating of 5.0/5.0, based on 10320+ user feedback. It should be noted that these data come from official self-declaration and have not been independently verified by a third party.

Industry Benchmarking: Alternatives to the traditional interview prep market are manual interview coaching ($200-$500/hour) and LeetCode membership (approximately $35/month). Interview Solver is positioned in the narrower scenario of "real-time interview assistance" - not to help users learn, but to provide real-time problem-solving support during the interview. There are certain ethical controversies in this category, and some companies explicitly prohibit the use of external aids in interviews.

Enterprise adoption: Undisclosed enterprise customer or B-side access cases. The product currently targets individual job seekers as its main customer base.

Cost Advantages of Interview Solver

Cost advantage: $39/month instead of $200-$500/hour for manual interview coaching

Interview Solver's pricing strategy revolves around "interview opportunity cost" - the official directly compares tool subscription fees with FAANG-level salary data (Amazon $359,703, Microsoft $402,000, Meta $548,909), emphasizing that "the cost of missing an offer far exceeds the tool subscription fee."

C client/individual users:

Plan	Price	Core Limitations	Applicable Scenarios
Free Trial	$0	10 messages, no credit card required	Experience functionality and compatibility testing
Monthly subscription (Monthly)	$39/month	Unlimited use, all features unlocked	Interview sprint period (1-3 months)

• The Free Truth: The 10-message trial credit is only enough to test tool compatibility and basic functionality, not enough to complete a full mock interview. Real usage scenarios require a paid subscription. The official mark is clearly "⚡ Limited to 10 messages during trial".
• Hidden Benefits: Compared to traditional interview coaching that costs $200-$500/hour, $39/month is equivalent to 1 hour of coaching in exchange for unlimited use for a full month. For candidates with a preparation period of 1-3 months, the total cost is $39-$117, which is significantly less than the $2,000-$10,000+ of traditional coaching.
• Hidden Cost: Over-reliance on real-time assistance may lead to "can do but not tell" in the interview - the interviewer not only evaluates the final code, but also evaluates the problem-breaking process and communication skills. If the tool response is delayed by more than 2-3 seconds, it can create noticeable pauses in live interviews.

API/Developer: No independent API interface is exposed. The product does not have an API calling mode for developers, and currently only provides desktop client form.

Enterprise/Private: No enterprise or private deployment options disclosed. The current stage of the product focuses on the individual user market.

Competitive product price comparison:

Products	Monthly Fees	Product Form	Core Differences
Interview Solver	$39	Native desktop client	Invisible screen sharing, LeetCode full question bank
Final Round AI	$149	Browser extension	Fewer features, 3.8x more expensive
UltraCode	Undisclosed	Desktop application	One of the competing products, lacks public pricing
Interview Coder	Undisclosed	Desktop application	One of the competing products, lacks public pricing
LockedIn AI	Undisclosed	Browser extension	One of the competitors, lacks public pricing

Main functions of Interview Solver

The functional system of Interview Solver revolves around the core scenario of "real-time interview assistance", and all functions are designed based on the premise of undetectability during the interview.

• Invisible Mode: core selling point. The application is completely invisible during screen sharing, supports transparency adjustment, and the window can be moved using Cmd+arrow keys. Invisible to Zoom ≤6.16 and all browser-based screen recording software. Applicable scenarios: Check the problem solutions provided by AI during real-time interviews without being discovered by the interviewer.
• Screengrabs: Send a screenshot of the contents of any window to AI, and AI will identify the question and return the solution. Supports all LeetCode question types. Applicable scenarios: When you encounter a problem you have never seen before, you can get the complete solution by taking a screenshot.
• Voice Transcription: Transcribe the interviewer's questions and the user's own answers at the same time, and AI automatically generates relevant answers and suggestions based on the conversation context. Applicable scenarios: In system design interviews or behavioral interviews, AI can capture questions in real time and generate corresponding answers.
• Companion Mode: Allows you to use any device (mobile phone, tablet) as a second screen to display codes and chat results. You can also invite friends or family to assist remotely. Applicable scenario: When the interviewer asks to share the home screen, view the AI ​​output through the companion device.
• Undetectable Hotkeys: Global hotkey control system, which can call out AI help when any application window is focused, without Alt+Tab switching windows. Applicable scenarios: Stay focused on the interview window and get prompts through hotkeys without causing suspicious operations.
• Document Upload: Upload your resume, project documents or technical information. During the interview, AI can give customized answers based on these personal documents. Applicable scenarios: The resume is carefully researched, and AI generates answers based on personal project experience.
• Invisible in Activity Monitor: Invisible in macOS Activity Monitor and Windows Task Manager. Applicable scenarios: Interviewer requirements

It is not found when viewing running processes.

• Flowcharts & Full Syntax Highlighting: Supports AI generation of system design flow charts and code solution display with syntax highlighting. Applicable scenarios: Draw architecture diagrams during system design interviews, and display professional code formats during coding interviews.

Expert View: The functional combination of Interview Solver is not isolated, but forms a collaborative link. Take a typical coding interview as an example - the interviewer dictates the question → Voice transcription automatically captures the question text → AI analyzes the question → Hotkey calls out prompts → Invisible floating window displays problem-solving ideas → Companion device synchronizes code solution. The entire process does not require any manual input or window switching. The interviewer can only see the candidate writing in the coding window. This "full-link invisibility" is its real competitive barrier, and it is far more than the strength of a certain function point.

Model and version evolution of Interview Solver

Interview Solver is currently in the early stages of product development, and the official has not disclosed a detailed version number system or update log. The following information is based on product milestones traceable on public pages:

Product development context (inferred based on public pages):

Time nodes	Stages	Key changes
~2026-07 (current)	1.0 initial version	Desktop application is online, supporting incognito mode, screenshot problem solving, voice transcription, companion mode, global hotkeys, document upload
Continuously iterating	Functions are perfected	The official page shows that compatibility continues to be updated (Zoom, Teams, Google Meet), and the question bank continues to expand

Version Features:

• The initial version is fully functional: Version 1.0 of Interview Solver has integrated 7 core functional modules, without adopting the "batched launch" strategy that was common in the early stages of the product. This implies that the product went through a long closed development cycle before release.
• Continuous coverage of the question bank: The official claims that the training data contains "entire set of all LeetCode questions", and the interview question bank covers mainstream technology companies such as Google, Amazon, Microsoft, Meta, Apple, and TikTok.
• Compatibility Maintenance: The official FAQ clearly states that it will be regularly updated to be compatible with the latest versions of Zoom, Teams, Google Meet and other platforms, indicating that the product requires continuous investment in adaptability maintenance.

Uncertainty: The official GitHub repository, version number mechanism or specific update frequency has not been disclosed. The actual version iteration speed of the product must be subject to official channels.

Technical advantages of Interview Solver

Interview Solver's technical advantages are concentrated in the three levels of "invisibility mechanism + native desktop architecture + vertical field optimization". The three constitute a combination barrier that is difficult to copy.

Native desktop architecture vs. browser extensions:

Interview Solver's choice of a native desktop app rather than a browser extension was a conscious technical decision. When a browser extension is screen shared, its tab or extension icon is visible to the interviewer; even if the extension window is hidden, traces of the extension running in the browser process can still be detected through the task manager. Native desktop applications can deeply call the window management API at the operating system level to achieve true "invisibility" - the window is not registered in the browser's rendering process, and the activity monitor and task manager cannot directly detect it.

Technical mechanism of invisible screen sharing:

The official description "Invisible to Zoom (≤6.16) and any browser-based screen recording software" implies that it takes advantage of a specific version of the screen capture API feature. On Windows and macOS, screen recording software typically captures frame content through low-level APIs such as DXGI / CGDisplayStream. Implementation of Interview Solver may involve:

• Utilize window cascading and transparency mechanisms to prevent application windows from being captured at the framebuffer level
• Or through the extremely fast display and concealment driven by hot keys, information display can be completed within the residual time of human vision.

Vertical domain optimized AI model:

The official claim is that the training data contains "complete LeetCode question bank". This means that its underlying model has been fine-tuned specifically for programming interview scenarios, rather than directly using zero-sample inference of a general large language model. The direct effect of this vertical optimization is:

• Understand the intention of algorithm questions more accurately (know whether the interviewer expects a "double pointer" or a "dynamic programming" route)
• The output format is more suitable for interview scenarios (with syntax highlighting, flow charts, and step-by-step explanations)
• The response latency is lower than the general model (the model size is optimized by tailoring)

Global Hotkey System:

By registering operating system-level global hotkeys (Global Hotkeys), Interview Solver can capture keyboard events in any window focus state without the need to switch window focus. This is crucial in an interview scenario - any Alt+Tab or mouse click to switch windows will reveal the fact that the candidate is looking at something else in the screen share.

Expert view: The real technical moat of Interview Solver is not in the AI ​​model itself (the model can be chased), but in the combined system engineering of "desktop invisibility + global hotkeys + real-time problem solving". Competing browser extensions have a ceiling in the invisibility dimension because browser sandboxes limit their access to the underlying operating system. This means that even if competing products are equal in AI capabilities, they are still at a disadvantage in the actual scenario of "use in interviews".

How to use Interview Solver

Step 1: Download and Installation: Visit interviewsolver.com, click "Get Started For Free" to register an account, download the desktop client and install it. No credit card required to try.

Step 2: Contextual configuration: After the installation is complete, configure global hot keys (the default shortcut keys can be used to call up the AI ​​floating window). It is recommended to complete the hotkey customization before the interview to ensure that it does not conflict with the shortcut keys of other applications.

Step 3: Preparation before the interview:

• Upload your resume and project documents (optional) so that AI can give customized answers based on your actual experience
• Select the target company (Google/Amazon/Meta/Microsoft/Apple/TikTok, etc.) and get the company’s previous interview questions
• Prepare behavioral interview materials in advance with the free ATS Resume Checker and STAR Method Builder

Step 4: Use during the interview:

• Launch Interview Solver when the interview starts (runs in the background, not shown in Activity Monitor)
• When the interviewer dictates the question, the voice transcription automatically captures the question
• Call up the AI floating window through global hotkeys to view problem-solving ideas and code solutions
• If the interviewer asks to share the screen, the AI solution automatically switches to companion device or invisible floating window mode
• Write the code solution provided by AI in the coding window (it is recommended to reorganize it in your own language to avoid completely consistent code templates)

Step 5: Review after the interview: Use the feedback mechanism of the tool to evaluate your performance and conduct special exercises to target weak areas.

Entrance control:

Entrance	Access method	Applicable scenarios
Desktop client	Official website download and installation	Real-time interview assistance (core entrance)
Companion mode	Access the Companion page from any device browser	Second screen when the home screen is shared
Free tool set	Official website /tools/ directory	Pre-interview resume optimization STAR preparation
Interview question bank	Official website /interview-questions/ Directory	Question practice by company classification

Interview Solver product pricing

Interview Solver's pricing system uses "interview opportunity cost" as its core narrative logic, rather than the functional tiered pricing of traditional SaaS. The median total salary for Senior Software Engineers at major technology companies (Amazon $359,703, Microsoft $402,000, Meta $548,909, etc.) is prominently displayed at the top of the pricing page to anchor users' expectations of "tool value."

Pricing Structure:

• Free Trial: $0, no credit card required. Includes full feature access, but limited to 10 messages. It is enough to verify the compatibility and basic effect of the tool on the user's system, but not enough to complete a complete interview assistance.
• Monthly: $39/month, no usage restrictions. It includes all functions such as incognito mode, screenshot problem solving, voice transcription, companion mode, global hotkeys, document upload, flow chart and syntax highlighting.

Free quota analysis: The free quota of 10 messages is too tight in terms of product experience. A typical coding interview may require 5-10 AI interactions (problem understanding, idea prompts, code optimization, complexity analysis), and the testing may be exhausted quickly. The free trial is positioned closer to "product validation" than "limited use."

Cost comparison with traditional interview coaching (deduction):

How to prepare	Monthly costs	Features	Hidden costs
Interview Solver	$39	Real-time assistance, unlimited times	May cause tool dependency
LeetCode Premium	~$35	Question bank + solution, no real-time assistance	Need to understand and practice by yourself
Manual interview coaching	$200-$500/hour	Highly personalized feedback	Total cost $2,000-$10,000+
Interview simulation platform	$50-$150/time	Live simulation, limited times	Risk of not being able to make an appointment at the right time

Purchase Suggestion: For job seekers with a preparation period of 1-3 months, you can first brush up on basic questions through LeetCode (combined with the free tools STAR Method Builder and Big O Cheat Sheet), and subscribe to Interview Solver 1-2 weeks before the interview for mock interviews and real-time auxiliary reinforcement in the sprint phase. This not only controls the cost ($39-$78), but also reduces the risk of tool dependence.

Application scenarios of Interview Solver

The core scenario of Interview Solver is "real-time problem-solving assistance" for technical interviews, and its value is most prominent in high-frequency, high-intensity, and highly competitive technical interview scenarios.

• FAANG Level Technical Interview Sprint: Interviews at Google, Amazon, Meta, Microsoft and other companies are famous for LeetCode Hard level questions and system design questions. Interview Solver's LeetCode full question bank coverage and system design flow chart generation capabilities directly correspond to the core pain points of this type of interview. Scenario Benefit: From "I have studied 200 questions but still panic when I take the exam that I have never seen before" → "Even if I encounter an unseen question, AI can immediately give the direction to solve the problem and output a executable solution on the spot."
• Safe Assistance for Screen Sharing Interviews: More and more technical interviews require candidates to share their screen and code in real time. Traditional browser extensions are completely useless in this scenario - switching tabs will be noticed by the interviewer. Interview Solver's Incognito Mode + Global Hotkeys + Companion Device combo is the only solution that assists in screen sharing scenarios. Scenario Benefits: From "No auxiliary tools can be used during screen sharing" → "The entire coding window is focused, and the AI ​​solution is seamlessly presented."
• Real-time architecture output for system design interviews: System design interviews require the completion of architecture design and drawings within a short period of time. Interview Solver can generate system design flow charts in real time and provide analysis of the pros and cons of architectural solutions. Scenario benefits: From "drawing architecture diagrams from memory" → "AI generates version evolution and component comparison in real time, and candidates make judgments and choices based on this" to demonstrate higher-level architectural decision-making capabilities.
• Personalized response generation for behavioral interviews: Combining uploaded resumes and project documents, AI can generate a STAR response framework based on personal experience for behavioral interview questions (such as "Tell me about a time..."). Scenario Benefits: From "temporary recall of project details" → "AI automatically extracts project highlights based on resumes and generates a structured answer framework".
• Real-time language assistance for non-native English speakers: For non-native English-speaking tech job seekers, Interview Solver’s voice transcription +

Real-time answer generation can help overcome language barriers and generate grammatically correct, professionally expressed English answers. Scenario Benefits: From "technical capabilities are underestimated due to poor language expression" → "AI-assisted generation of authentic expressions, focusing on technical content output".

Quantitative deduction of cost reduction and efficiency improvement: Take a software engineer preparing for Meta E5 interview as an example - the traditional preparation path: 3 months of LeetCode test questions ($35/month × 3 = $105) + 5 times of mock interview coaching ($200/time × 5 = $1,000) + special preparation for system design ($500) ≈ $1,605. Alternative path with Interview Solver: 3 months of LeetCode + final 1 month of Interview Solver subscription ($39 × 1 = $39) + 1 live mock interview verification ($200) ≈ $274. Cost reduction of approximately 83%. This is a deduction, and actual results will vary depending on individual bases and target companies.

Human-machine collaboration boundary:

• Can be 100% automated: LeetCode question analysis, code solution generation, system design flow chart drawing, behavioral interview framework generation, voice transcription and question understanding.
• Manual confirmation required: The final submitted code solution (must ensure that the logic is correct and meets the interviewer's expectations), the technical selection judgment of system design (AI can propose options, but the candidate's decision-making and explanation are the key to scoring), and the compliance judgment of tool use (the interview policy of the target company needs to be confirmed).

Interview Solver is suitable for people

The positioning of Interview Solver determines that it has a clear boundary of "perfect fit" and an equally clear boundary of "absolutely not applicable".

• Software Engineer (SDE 1-3 / Senior+): Interview Solver’s core user group. Whether you are preparing for Amazon's Leadership Principles interview, Google's algorithm interview, or Meta's system design interview, the tool's full question bank coverage and real-time assistance capabilities directly respond to your needs. Deduction: An SDE 2 candidate used Interview Solver to pass Microsoft's phone interview (from official user testimonials), and it took about 2-4 weeks of using the tool to produce obvious results.
• Current students majoring in computer science (junior year/second year graduate student/fresh graduate): Campus recruitment interviews are highly competitive, and candidates need to cover as many topic types as possible within a limited time. Interview Solver's LeetCode full question bank and real-time problem-solving assistance can help quickly fill in knowledge blind spots. Deduction: From "Brushing the basics of 50 questions" → "AI assistance covers 200+ question types, and there are strategies to deal with questions that have never been done before during the interview."
• Career changers (non-CS background turning to software development): Job seekers who teach themselves programming usually lack systematic training in data structures and algorithms, and their weak foundation is easily exposed in interviews. Interview Solver can bridge this gap during interviews. Deduction: From "self-study for 6 months but came up blank when I shredded the code in the interview" → "AI prompts problem-solving ideas in real time, returning the focus to expression and communication".
• Non-English native technical job seekers: In English interviews, language barriers may lead to technical abilities being underestimated. Speech transcription + real-time answer generation can help overcome this obstacle. Deduction: From "being judged to have insufficient communication skills due to poor expression" → "AI assists in generating smooth technical answers, focusing on demonstrating technical depth".

Does not fit boundaries:

• Companies with zero tolerance for interview aids: Some technology companies (especially in the financial technology and cybersecurity fields) explicitly prohibit the use of any external aids in their interview policies. Using Interview Solver may violate the interview agreement, resulting in the offer being revoked or even being blacklisted in the industry. Recommendation: Be sure to confirm the interview policy of the target company and assess the risks before using it.
• Job seekers seeking authentic competency assessments: If the goal is to learn through interviews rather than just interviewing, over-reliance on AI assistance can undermine real competency improvement. Interview Solver is more suitable as a "weakness-filling tool" rather than a "daily dependency".
• Non-technical job seekers: Interview requirements for non-coding positions such as product managers, designers, and sales are not within the core coverage of Interview Solver (although there are free tools such as STAR Method Builder to assist).
• Senior Architect/Principal Engineer Level: The interview focus at this level is no longer on LeetCode problem solving, but on system design depth, technical strategy and leadership assessment. Interview Solver is of limited assistive value in such scenarios.

Summary and Outlook of Interview Solver

Core Competencies: Interview Solver has established a differentiated positioning in the AI interview assistance track with the combination of "invisible desktop client + LeetCode full question bank AI + global hotkeys + companion mode". Compared with competing browser extension products, its native desktop architecture has an insurmountable advantage in invisibility. The pricing of $39/month has a clear cost-effective advantage compared to Final Round AI's $149/month. Supplemented by the community base of "Trusted by 10428+ engineers" and a user rating of 5.0/5.0, the product has already established a certain brand trust in the early stages.

Current Limitations:

• The product is in the 1.0 stage, has a short version history, and long-term stability and compatibility maintenance have not been verified over time.
• The official has not disclosed the specific information of the underlying AI model (model size, training data range, update frequency), and the technical debt and model iteration route are not transparent.
• Interview assistance tools face continued ethical controversy and platform confrontation (updates of Zoom/Teams/Google Meet may break the invisibility function at any time)
• Lack of APIs and enterprise-level solutions, the business model relies solely on personal subscriptions
• The free trial limit of 10 messages is too tight and user decision-making costs are high.

Industry trend judgment:

• The penetration rate of AI-assisted tools in technical interviews will continue to increase, and "using AI in interviews" will evolve from fringe behavior to industry norm. It is expected that within 1-2 years, the interview process may usher in structural changes (such as mandatory offline coding or full AI proctoring), which will have a positive or negative impact on Interview Solver's business model.
• Competing browser extension products may be further restricted by browser security policy updates (such as Manifest V3), which favors Interview Solver in the form of native desktop applications.
• FAANG companies may increase investment in interview anti-cheating (keyboard logging, window monitoring AI behavior analysis), forcing Interview Solver to continue to upgrade stealth technology.

Procurement/Adoption Risk Assessment:

• Technical Risk: Incognito mode works for Zoom ≤6.16, but may not work after Zoom is updated to 7.0. It is recommended to test compatibility on your own device before subscribing, and check the target interview platform version before the interview. The compatibility information disclosed on the official real-time page shall prevail.
• Compliance Risk: Using AI to assist in technical interviews may violate some companies’ interview integrity policies. Users are advised to: ① Check the interview policy document of the target company ② Assess the risk of detection (the official claim is that "no user has been detected during use" but absolute safety cannot be guaranteed) ③ After signing the offer, weigh whether it is necessary to disclose usage behavior to the employer. For industries with strict compliance requirements such as financial technology and network security, it is recommended to use it with caution.
• Capability Risk: Long-term reliance on real-time assistance may erode independent problem-solving abilities. It is recommended that Interview Solver be positioned as a "short board tool for the interview sprint stage" rather than a "substitute for the question brushing stage" - first use LeetCode to establish basic capabilities, and then use the tool for simulation and auxiliary reinforcement 1-2 weeks before the interview.
• Alternatives: If Interview Solver is not available in your area or the compliance risk is too high, consider LeetCode Premium (pure question bank, no real-time assistance) or human interview coaching (high cost but compliant). The former is $35/month and the latter is $200-$500/hour. When making a decision, you need to comprehensively evaluate the budget, interview urgency, and the compliance requirements of the target company.

Related tools: notion-ai, google-workspace

Version Info

• Interview Solver ：There is no official precise date yet. Desktop application form provides core functions such as incognito mode, screenshot problem solving, voice transcription, companion mode, global hotkeys, and document upload. （~2026-07）
• Interview Solver ：There is no official precise date yet. The first public version, positioning AI interviews the co-pilot. （~2026-07）