Practice should adapt, evaluate, and explain.
The platform brings interview preparation into one structured product experience rather than treating questions, resumes, coding tasks, and progress as disconnected tools. Users can configure a target role, interview type, difficulty, question count, and optional resume context before entering an AI-guided mock interview.
The completed system uses a hybrid architecture: Django owns the core application, account history, resume workflow, and gamification, while a FastAPI service handles the migrated interview AI slice through local Ollama.
Different product areas follow the path that fits them.
Hybrid application flow
Configuration leads to structured feedback.
Question generation
Builds an interview around the selected role, type, difficulty, question count, and optional resume context.
Answer evaluation
Returns a score, verdict, feedback, strengths, improvements, and missed points for review.
Interview reporting
Converts individual evaluations into a structured report that supports reflection after the session.
Fallback behavior
Can use deterministic/local fallback behavior if the Ollama or FastAPI interview service is unavailable.
Preparation extends beyond the interview session.
The resume workflow validates uploads, parses text, scores the material, and returns strengths, weaknesses, and suggestions. This workflow currently belongs to the Django-side AI path rather than the migrated FastAPI interview slice.
Gamified practice covers technical quizzes, problem solving, and behavioral scenarios. Progress, XP, levels, streaks, badges, and a leaderboard provide continuity, while coding challenges can execute inside restricted Docker environments.
The hybrid boundary reflects the product’s real evolution.
Django core
Accounts, history, resume processing, and gamification remain part of the primary application.
FastAPI interview slice
The interview AI client reaches a dedicated service for question generation and evaluation.
Local inference
Ollama and Mistral support the interview service without implying every AI feature uses that same path.
Resilient behavior
Fallback logic keeps interview flows usable when the local AI service is unavailable.
A product stack built around clear responsibilities.
Application
Django · Django REST Framework · PostgreSQL
AI service
FastAPI · Ollama · Mistral · Local fallback behavior
Assessment
Mock interviews · Resume analysis · Structured reports
Practice
Quizzes · Scenarios · Progression · Docker-isolated coding