Analytical benchmark
Deterministic scenarios cover metadata inflation from 1× through 16×. They are explicitly not production measurements.
Python 3.12+IEEE companionPost-quantumMIT
Quantum-safe lakehouse research
Keep cryptographic growth
off the planning hot path.
AQPO studies predictive caching, verified segment materialization, and compact metadata views for open table layouts under post-quantum overhead.
15curated references
LOWfinal AEGIS risk
0flagged citations
10artifact tests
Research checks and Pages deployment passing
Architecture
A verifiable metadata acceleration layer
AQPO sits between compute engines and authoritative catalogs. Optimization never replaces the source of truth.
01Query enginesTrino · Spark
→
02AQPO proxyPredict · verify · compact
→
03CatalogREST · snapshots
→
04Object storageAuthoritative metadata
Research ledger
Inspectable assumptions, bounded claims
Trace-replay regression
Thirty fixed seeds increased mean hit ratio from 0.706375 to 0.728542, while 0.054240 precision exposed a deployment limitation.
Deny-by-default gate
Manuscripts, raw scans, local paths, private email, and unauthorized attribution fail publication.
Modeled 16× metadata read66.2 → 23.8 MBBaseline-PQ versus AQPO-Agentic
Illustrative analytical output. Executed trace replay is reported separately; production validation remains future work.
Integrity
Transparent detector status
Citation integrityCompleted · 0 flagged issues
AI-content signalCompleted · 0.14 · probabilistic
External matchingUnavailable · no corpus loaded
Self-overlapUnavailable · no verified prior works
The 2026-09-02 scan reported LOW risk, 0% citation issues, grammar quality 0.99, and IEEE compliance. Detector outputs are probabilistic indicators, not proof of authorship or misconduct. Raw reports remain private because they contain manuscript-derived text.
Reproduce
Inspect the model locally
The public repository contains sanitized analytical inputs, validation code, and a deny-by-default privacy check. It does not contain the manuscript or raw scan reports.
quantalake-aqpo
git clone https://github.com/darisisuresh/QuantaLake-AQPO.git
cd QuantaLake-AQPO
python3 -m unittest discover -s tests -v
python3 scripts/validate_evidence.py
python3 scripts/privacy_check.py
Research stewardship
Suresh kumar Darisi
Sole repository author and contributor. The public project contains reproducibility artifacts and documentation; the manuscript remains private.