Mock AI
Generates and grades interviews with an LLM while keeping auth, sessions, and scoring accountable to a real backend.
project overview
MockAI is a full-stack mock interview and assessment platform that generates technical tests, simulates live interviews, and evaluates responses using Google Gemini. The platform combines a FastAPI backend with a React frontend to support AI-generated assessments, conversational onboarding, voice-enabled interview sessions, realtime feedback, and persistent performance tracking through MongoDB-backed user accounts and JWT-secured sessions.
architecture
The React frontend handles onboarding flows, test-taking interfaces, dashboards, and browser-based voice interactions using Web Speech APIs. The FastAPI backend manages authentication, session handling, question generation, answer evaluation, and persistence services. Gemini 2.0 Flash powers both test generation and qualitative answer evaluation through backend-controlled API routes. MongoDB stores users, sessions, generated assessments, and evaluation history using Motor with pooled async connections and unique-email enforcement. The platform supports conversational test creation, role-specific interview simulations, timed prompts, speech synthesis, speech-to-text capture, and detailed feedback dashboards.
constraints
- LLM accountability — generated questions and feedback flow through one backend service, never straight from the browser to the model
- session integrity — assessments mean nothing if identity and history are loose; JWT and unique-email enforcement are load-bearing
- voice latency — the capture → evaluation → synthesis round-trip must stay conversational without any server-side audio infrastructure to tune
tradeoffs
- browser Web Speech APIs over a server-side ASR pipeline: zero audio infrastructure, at the cost of device-dependent recognition quality
- Gemini for both generation and evaluation over separate graders: one model contract, consistent rubric, single point of prompt discipline
failure notes
- a reserved java-backend module sits in the repository, currently empty — an honest placeholder for planned integration rather than a hidden dead end
infrastructure
fastapi · react 19 · mongodb · gemini 2.0 flash · web speech api · jwt auth
engineering reasoning
Designed to keep authentication, session management, question generation, and evaluation inside a centralized backend service instead of exposing model interactions directly to the client. Browser-native voice APIs were used to enable realtime interview flows without introducing dedicated audio-processing infrastructure.
future work
- > the reserved Java service integration
- > richer rubric-anchored scoring beyond qualitative feedback