RAHUL ANANTMathematics & Computing
Back to Projects Library
AI / MLCompleted Prototype2024

AI Interview Preparation Mentor & Real-Time Evaluator

Role: Lead AI Engineer

Full-stack AI mentorship platform conducting real-time technical interviews with automated code evaluation, semantic hint generation, and performance analytics.

AI Interview Preparation Mentor & Real-Time Evaluator

Technology Stack & Libraries

TypeScriptPythonLangChainOpenAI APINext.jsTailwind CSSDocker
The Challenge / Problem

Context & Objectives

Job seekers lack access to realistic, unbiased mock interviews that provide rigorous architectural feedback without paying prohibitive coaching fees.

Architectural Solution

Engineered Approach

Designed an asynchronous evaluation loop combining AST code parsing with conversational LLM feedback agents that guide candidates through problem solving.

System Architecture & Pipeline

Next.js Frontend -> WebSocket Stream -> LangChain Orchestration -> Dockerized Pyodide Sandbox -> Evaluation Vector Store.

Engineered an adaptive technical mock-interview system that replicates elite engineering interview loops. The platform leverages LLMs with strict system prompting and sandboxed code execution to evaluate algorithmic correctness, time complexity, and candidate explanation clarity.

Engineered Capabilities & Innovations

Multi-turn conversational interview agent with dynamic difficulty adaptation
In-browser sandboxed code execution for Python and JavaScript
Detailed complexity analysis report identifying suboptimal loop constructs
Actionable improvement roadmaps generated after each mock session

Empirical Results & Benchmarks

Conducted 250+ simulated technical interviews during university pilot testing, reducing candidate preparation anxiety and improving problem articulation.

Have Questions About This System?

I am always glad to discuss technical architecture, benchmarks, or potential collaboration.