Submission: Intelligent Resume Assistant (HireMind AI)

Hello team,

I am submitting my implementation for the Intelligent Resume Assistant (Agentic AI System) assignment.

I built HireMind AI—a production-style MVP focused on reliable, controlled agentic behavior rather than a simple LLM wrapper. To prevent hallucinations and enforce strict compliance with the required JSON guardrail contract, the system implements a "deterministic-first" routing approach, using Python-based parsers and keyword extractors for explicit resume data, while reserving LLM capabilities (via Gemini) exclusively as a fallback for complex reasoning and synthesis queries.

Implementation Highlights:
Architecture: Next.js (Frontend) + FastAPI (Backend) with an in-memory session manager.
Agentic Design: A custom orchestrator that evaluates intent, selects the correct tools, and applies strict Pydantic guardrails before responding.
Transparency: Built an "Agent Trace" panel in the UI so recruiters can audit exactly how the AI arrived at its conclusion, including confidence scores and extracted evidence.
Missing Data Handling: Fully compliant with the requirement to explicitly state when evidence is missing rather than guessing.
Links:
Repository Link: https://github.com/Maviya13/Intelligent-Resume-Assistant-.git
Demo / Walkthrough: Full local setup instructions are provided in the repository's README.
Thank you for the opportunity, and I look forward to your feedback!

Best regards,
Maviya Mahagami