AnchorCV – Context-Aware Mock Interview

Most AI interview tools check if you match the job. This one checks if you can defend your own resum, the one document where every claim came from you, not a recruiter.

Comment

Overall Concept & Goal of AnchorCV

AnchorCV is an AI-powered career platform built primarily for international students and job seekers navigating unfamiliar job markets abroad. Its broader goal is to replace the scattered, manual process of job hunting — tailoring a resume by hand for every role, guessing at ATS keyword gaps, writing cover letters and outreach emails from scratch, and prepping for interviews with generic advice — with one connected workflow: analyze a resume and a target role once, then carry that understanding through resume building, ATS scoring, job matching, and interview preparation, instead of starting over at every step.

The Context-Aware Mock Interview is the newest piece of that platform, built during OpenAI Build Week, and it extends the same core idea — understand the candidate's actual background and the actual role, not a generic template — into interview preparation specifically.

Inspiration

Preparing for interviews is frustrating because most AI interview platforms ask the same generic questions to every candidate. They rarely understand a candidate's background, previous experience, or the specific requirements of the role.

We wanted to build an interview experience that feels much closer to speaking with a real interviewer. Instead of simply matching keywords, AnchorCV analyzes both the candidate's resume and the target job description to understand the candidate's experience, required skills, and expectations of the role before planning the interview.

More importantly, a real interviewer doesn't immediately score an incomplete answer—they ask follow-up questions. That became the core idea behind this project.

What it does

AnchorCV's Context-Aware Mock Interview creates personalized interview sessions based on both the candidate's resume and the target job description.

The interview workflow includes:

AI-powered resume and job description analysis Interview planning before questions are generated Personalized interview questions tailored to the candidate Context-aware clarification questions when additional evidence is needed Evidence-based answer evaluation A comprehensive Interview Readiness Report

Instead of immediately evaluating a short or vague answer, the AI first determines whether enough evidence has been provided. If not, it asks an adaptive clarification question before making its assessment.

At the end of the interview, candidates receive detailed feedback including communication analysis, relevance, strengths, weaker areas, interview summaries, and personalized coaching recommendations.

How we built it

We built the feature on top of the existing AnchorCV platform using:

React FastAPI Supabase OpenAI API Whisper for voice transcription

The interview pipeline was redesigned into multiple stages:

Resume and job description understanding AI interview planning Personalized question generation Clarification and evidence collection Evidence evaluation Interview Readiness Report generation

Throughout OpenAI Build Week we used OpenAI Codex extensively to accelerate development, implement the interview workflow, refactor existing architecture, improve the evidence evaluation pipeline, debug production issues, and rapidly iterate on new features while maintaining a stable codebase.

Challenges we ran into

The biggest challenge was designing an interview that behaves more like a real interviewer instead of a simple question generator.

Building the clarification stage required changing the evaluation flow so the AI could first decide whether enough information had been provided before assigning a score.

Another challenge was integrating these capabilities into an existing production application without introducing duplicate workflows or sacrificing performance.

Accomplishments that we're proud of

Our biggest accomplishment is creating an interview experience that adapts to the candidate instead of treating every applicant the same.

We're especially proud of:

Context-aware interview planning Adaptive clarification questions Evidence-based evaluation instead of immediate scoring A comprehensive Interview Readiness Report Successfully integrating the feature into an existing production platform during Build Week What we learned

This project reinforced that building useful AI applications is about much more than calling an LLM.

Creating a realistic interview experience required careful workflow design, context management, structured evaluation, and thoughtful user interaction.

We also learned how valuable OpenAI Codex can be for rapidly implementing, debugging, and refining complex application features during a short development cycle.

What's next for AnchorCV – Context-Aware Mock Interview

Our roadmap includes:

Voice-first interview sessions Multi-round interviews Company-specific interview styles Team and recruiter dashboards Long-term interview progress tracking Interview analytics across multiple sessions Resume improvement suggestions based on interview performance

Our long-term vision is to make interview preparation as personalized and realistic as the interviews candidates will face in the real world.

Built With

Share this project:

Updates