{"paper_id":"3b7ec14d-3a5f-4a5a-b139-ebe5906e82db","body_text":"AEVUM: An Agent-Native Persistent Memory Database System with Autonomous Data Management and Multi-Stage Compression | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 23 February 2026 V1 Latest version Share on AEVUM: An Agent-Native Persistent Memory Database System with Autonomous Data Management and Multi-Stage Compression Author : Jalendar Reddy Maligireddy 0009-0002-5881-0084 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.177187991.15165347/v1 185 views 88 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Modern AI agent systems increasingly operate on data infrastructure designed for human users: rigid relational schemas, manually authored queries, and external access-control layers that sit entirely outside the agent's reasoning loop. This architectural mismatch imposes three compounding overheads on agentic software: schema migrations for every evolving field, manual glue code to translate LLM-generated intents into structured queries, and fragmented infrastructure that separates structured records, vector memory, and audit logs into uncoordinated silos. We present AEVUM, an agent-native embedded database system that inverts this model. In AEVUM the AI agent owns, manages, and queries persistent storage autonomously through four core abstractions: (1) schema-free spaces whose column layout is inferred at flush time, (2) an immutable agent-identity manifest embedded in every database file for provenance and cross-agent trust, (3) lazy cross-database reference federation that links multiple .aevum files without ETL copies, and (4) a provideragnostic natural language query layer that dispatches over OpenAI, Anthropic, Gemini, Ollama, and any OpenAI-compatible endpoint. Storage efficiency is achieved through a deterministic five-stage compression pipeline: columnar transposition, type-aware binary encoding (delta integers, float32 packing, bit-packed booleans, dictionary strings), content-addressed block deduplication via BLAKE3, entropy pre-processing (zero-run encoding, byte transposition), and selectable C-level codecs (Zstandard, LZ4, Brotli). On five public real-world datasets AEVUM achieves 5-16× compression over raw JSON and outperforms gzip-9 by 10-25% at comparable ratios. NL query latency ranges from 95 ms (local Ollama, simple recall, p50) to 620 ms (remote provider, cross-database join, p95); the storage and retrieval layer itself contributes less than 5 ms. Private databases are protected by ChaCha20-Poly1305 authenticated encryption with PBKDF2-HMAC-SHA256 key derivation (480,000 iterations). The complete system deploys as a single Python wheel with no external server dependency, making it suitable for resource-constrained edge and embedded agent deployments. Supplementary Material File (aevum_jalendar.pdf) Download 741.32 KB Information & Authors Information Version history V1 Version 1 23 February 2026 Copyright This work is licensed under a Creative Commons Attribution 4.0 International License Keywords aevum agent-native database chacha20-poly1305 columnar compression embedded storage natural language query persistent agent memory schema-free storage Authors Affiliations Jalendar Reddy Maligireddy 0009-0002-5881-0084 [email protected] View all articles by this author Metrics & Citations Metrics Article Usage 185 views 88 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Jalendar Reddy Maligireddy. AEVUM: An Agent-Native Persistent Memory Database System with Autonomous Data Management and Multi-Stage Compression. Authorea . 23 February 2026. DOI: https://doi.org/10.22541/au.177187991.15165347/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(\".js__slcInclude\").on(\"change\", function(e){ if ($(this).val() == 'refworks') $('#direct').prop(\"checked\", false); $('#direct').prop(\"disabled\", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {\"doi\":\"10.22541/au.177187991.15165347/v1\",\"type\":\"Article\"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob(\"bG9jYXRpb24=\"),_bnb=atob(\"b3JpZ2lu\"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(\" \")); $.get(\"/resource/lodash?t=\"+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML=\"window.__CF$cv$params={r:'9fe5320d6d8b58d3',t:'MTc3OTIxNTY2NA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);\";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();","source_license":"CC-BY-4.0","license_restricted":false}