Memora AI — Free Your Cloud. Rediscover Your Life.
Tagline: Your private AI memory vault that transforms scattered photos, videos, and files into searchable, connected memories.
Inspiration
We capture thousands of photos and videos throughout our lives, but finding a specific memory years later can be surprisingly difficult.
At the same time, Gmail, Google Drive, OneDrive, and Google Photos storage keeps filling up. Users are forced to choose between purchasing more cloud storage and deleting valuable content.
We wanted to solve both problems.
What if your NAS could not only preserve your digital memories, but actually understand how they connect?
What It Does
Memora AI transforms UGREEN AI NAS into an intelligent, privacy-focused memory vault.
It is designed to help users:
- Archive selected cloud files, photos, and videos onto their NAS.
- Identify duplicate content and potential cloud storage savings.
- Organize media using AI-generated descriptions and metadata.
- Build a multimodal memory knowledge graph connecting events, places, people, dates, and media.
- Search photos and videos using natural-language descriptions.
- Retrieve relevant moments inside longer videos.
- Verify archived content before recommending cloud cleanup.
For example, a user could ask:
"Find the video where my father was laughing during my sister's birthday."
Memora would search available visual descriptions, transcripts, timestamps, and event relationships to identify relevant media.
How It Works
Memora uses a proposed Multimodal Graph RAG workflow.
Step 1 — Import: The user selects authorized media stored on UGREEN NAS or imports files from supported cloud exports.
Step 2 — Understand: Supported AI and knowledge capabilities extract metadata, visual descriptions, and contextual information.
Step 3 — Connect: Memora organizes relationships between events, dates, locations, and media into a memory graph.
Step 4 — Retrieve: Natural-language queries combine semantic matching, metadata filtering, and graph relationships.
Step 5 — Preserve: Archive verification helps users understand whether important media remains recoverable.
The workflow is designed to prioritize local NAS processing, user authorization, and evidence-backed retrieval.
What Makes It Different
Traditional cloud storage systems focus on storing files.
Traditional photo search focuses on identifying objects, faces, or locations.
Memora focuses on connected human memories.
Instead of searching for "cake," users can describe an entire experience involving a person, event, approximate date, and action.
Memora aims to use those relationships to retrieve relevant photos and video moments.
Its second differentiator is connecting intelligent memory search with safe cloud archival.
Challenges
The main technical challenges are multimodal indexing, event relationship extraction, accurate video timestamp retrieval, and maintaining data privacy.
Another challenge is implementing these workflows within the officially supported UGREEN Skill framework and available expert-agent capabilities.
Cloud integrations also depend on provider API permissions and restrictions.
Future Vision
We envision Memora becoming a personal memory intelligence layer for NAS storage.
Future versions could support richer cloud integrations, family memory collections, conversational memory exploration, automated archival recommendations, and recovery-readiness checks.
Our mission is simple: Your memories should belong to you, remain protected, and be as easy to find as they are to remember.
Built With
- ai
- markdown
- rag
- ugreen-ai-nas
- uliya-skill-framework