InterviewOS is an AI-powered technical interview platform designed to simulate a structured, adaptive technical interview.
Built for the ABTalks AI Engineering Hackathon, InterviewOS evaluates candidates based on their answers, adapts the next question according to performance, tracks knowledge gaps, stores interview memory using Breeth, and generates a final performance report.
InterviewOS does not follow a completely fixed question sequence.
The next question is influenced by the candidate's previous performance:
- Low score → revisit the current topic
- Medium score → move to the next topic
- High score → move forward with deeper questions
This creates a more realistic interview experience.
The interview is generated based on candidate information such as:
- Name
- Job role
- Years of experience
- Education
- Interview objectives
- Curriculum topics
The system creates an interview plan containing multiple technical objectives.
InterviewOS generates questions according to the selected technical topic.
Supported areas include concepts such as:
- Python
- Programming Fundamentals
- Data Structures & Algorithms
- Object-Oriented Programming
- SQL
- Database Systems
- Machine Learning
- Deep Learning
- Generative AI
- Large Language Models
- APIs & Backend Development
- FastAPI
- Cloud Computing
Candidate answers are evaluated using multiple signals:
- Answer depth
- Answer length
- Technical terminology
- Topic relevance
- Technical concepts used
Each answer receives a score from 0–10.
The evaluator also identifies:
- Strengths
- Knowledge gaps
- Feedback
InterviewOS tracks candidate performance across interview topics.
Weak topics are identified and converted into personalized next steps.
Example:
Weak Topic
↓
Data Structures & Algorithms
↓
Knowledge Tracker
↓
Recommended Action
↓
Review Data Structures & Algorithms