Add Level-3 academic integrity roadmap and FastAPI scoring scaffold

[Suhaskumard]

Motivation

• Move the project toward an enterprise-grade Academic Integrity & Research Authenticity platform that supports idea-level detection, explainability, and institution-scale analytics.
• Provide a prioritized, implementation-ready roadmap that sequences governance, detection intelligence, research-grade features, and enterprise scaling.
• Ship a minimal, auditable backend scaffold to let downstream work integrate scoring, policy rules, and explainable evidence quickly.

Description

• Rewrote README.md to reflect the new Academic Integrity & Research Authenticity direction and published a prioritized Top 10 must-build list and usage notes for local runs with uvicorn and requirements.txt.
• Added a detailed blueprint at docs/academic-integrity-next-level.md outlining architecture layers, phased rollout, data entities, the proposed Academic Integrity Index formula, explainability contract, and suggested KPIs for research/IEEE readiness.
• Implemented a lightweight FastAPI scaffold at backend/main.py with typed Pydantic models (IntegrityPolicy, IntegritySignals, EvidenceItem, IntegrityScoreResponse) and endpoints: GET /health, POST /policies, GET /policies, and POST /score which computes a composite integrity index and returns evidence items.
• Added pinned backend dependencies in backend/requirements.txt for fastapi, uvicorn, and pydantic to support the scaffold and future integration of model adapters.

Testing

• Ran python -m py_compile backend/main.py to validate the Python syntax and it completed successfully.
• No unit/integration tests were added in this change; further CI and runtime tests (endpoint behavior, policy persistence, and model integrations) are planned as follow-ups.

---

Codex Task