Deployment of a self-hosted environment to ensure absolute data privacy and independence from cloud dependencies.
Embedding regulatory and institutional documents to establish an automated, precise reference oracle.
Connecting the internal LLM brain directly with reporting scripts, calculators, and document archiving.
Proposed extensions once the core RAG knowledge base and compliance workflows are stable.
A campus-facing chat interface so budget officers, accountants, and department heads can query circulars and manuals in plain language, with citations.
Extend the RAG layer beyond Budget Services to Accounting, HR, Registrar, and satellite campuses, on a shared but permissioned index.
COA-ready logging of every AI-assisted answer, quarterly circular refresh cycles, and human sign-off before any output feeds a report.
Your sovereign, campus-wide knowledge assistant for PUP's budget and finance offices.
P.R.I.N.C.E.S.S. runs entirely on university-controlled infrastructure — no query ever leaves campus. When you ask a question, it searches the relevant document collection first, then answers only from what it finds there, citing the source.
Air-gapped inference — nothing sent to third-party AI providers.
Every response is grounded in an actual circular, code, or manual.
The knowledge base is re-ingested as new circulars are issued.