HireMe

Overview

HireMe is an AI-assisted technical interview platform built during the SKIPS University BrahmaCode Hackathon.

The platform aims to make technical interviews more structured and data-driven by combining live coding assessments with behavioral signals such as facial expressions and candidate engagement.

The system allows interviewers to conduct real-time coding interviews while simultaneously monitoring candidate behavior, generating analytics, and evaluating coding performance.

The project was built in 36 hours by a team of five, resulting in a fully functional prototype that won 1st prize (₹50,000) at the hackathon.

Problem

Traditional technical interviews rely heavily on subjective evaluation and limited signals. Interviewers often struggle to assess:

• candidate engagement during interviews
• behavioral cues such as stress or confidence
• real-time coding ability under interview conditions
• consistent evaluation across multiple candidates

Additionally, many interview processes require separate tools for video calls, coding environments, and evaluation metrics.

The goal of HireMe was to build a single platform that combines live interviews, coding evaluation, and behavioral analytics to create a more transparent and structured hiring process.

Constraints

• Hackathon time limit – the entire system had to be designed and implemented within 36 hours.
• Real-time collaboration – the platform required live video, coding, and event synchronization.
• Low infrastructure complexity – solutions had to rely on existing APIs and lightweight architecture.
• Accuracy limitations – behavioral analysis needed to work with browser-based models.
• Team coordination – development had to be split across frontend, backend, and AI components.

Key Engineering Decisions

Real-time interview sessions using WebSockets

Socket.IO was used to synchronize interview events between interviewer and candidate including code updates, session state, and live interactions.

Reason:
Real-time communication is essential for collaborative coding interviews.

Tradeoff:
Managing connection state and synchronization across multiple clients adds complexity.

Embedded coding environment using Monaco Editor

The coding interface was built using Monaco Editor, the same editor used in VS Code.

Reason:
Provides a familiar development experience and supports syntax highlighting, multiple languages, and editor extensions.

Tradeoff:
Embedding a full-featured code editor increases frontend complexity.

Automated code evaluation using Judge0 API

Code submissions were evaluated using the Judge0 API to run candidate code in isolated environments.

Reason:
Building a secure code execution environment from scratch is complex and risky.

Tradeoff:
External API dependency for code execution.

Facial expression analysis using Face-api.js

Face-api.js was used to detect facial expressions and engagement signals from the candidate’s webcam during interviews.

Reason:
Provides browser-based machine learning models capable of running facial detection without requiring server-side processing.

Tradeoff:
Browser-based ML models may have accuracy limitations depending on lighting and camera quality.

Analytics dashboard for interview insights

Interview performance and behavioral signals were visualized using Recharts.

Reason:
Providing interview analytics helps interviewers evaluate candidates beyond coding performance.

Tradeoff:
Behavioral metrics should be interpreted carefully to avoid biased conclusions.

Results

• Built a functional end-to-end technical interview platform within 36 hours.
• Enabled real-time coding interviews with synchronized editing.
• Integrated webcam-based facial expression tracking during interviews.
• Implemented automated code evaluation through Judge0.
• Delivered performance analytics and candidate engagement insights.
• Won 1st prize (₹50,000) at the SKIPS University BrahmaCode Hackathon.

Takeaways

Building HireMe under hackathon constraints highlighted the importance of rapid architecture decisions and practical tradeoffs.

Key lessons:

• Leveraging existing APIs accelerates development under time constraints.
• Real-time collaboration systems require careful event synchronization.
• Browser-based machine learning can enable interesting product features without heavy infrastructure.
• Clear task separation across teams is essential when building complex systems in a short time.