AI that works. As aspected.

RAG Pipelines are becoming too complex

Vector databases optimize for similarity. Enterprise AI require context-aware relevance

Today’s RAG systems rely on layered fixes, metadata filters, rerankers, and graph logic because the retrieval layer lacks context.
Aspected introduces the Aspect Database: a new retrieval architecture that incorporates context directly into retrieval.

Early preview for RAG developers and AI architects

Start buildingSee the architecture
Aspect Database

Why RAG pipelines keep getting more complex

❗ Context is added after retrieval 
As similarity search lacks context, teams keep adding layers. This increases complexity and reduces reliability.

Vector Database Overview
Compare
Nothing else matters
Fixing Vector Database limitations
Compare
Nothing else matters
ADB
Compare
Nothing else matters

 

Why vector search breaks enterprise AI

Vector search retrieves similar text. But enterprise AI requires more than similarity. 

Two Aspected developers at a desk
Impressive work
Nothing else matters

AI systems must consider:

  • the correct version of a document

  • whether content is approved and valid

  • the role of the user

  • the current step in a workflow

  • compliance rules and constraints

Because this context is not part of retrieval, it is added afterwards through filters, rerankers, and orchestration.  This leads to complex and fragile systems.

Similarity is not enough. Retrieval needs context.

Vector Search vs. Aspect Search

Results generated from a production dataset of ~240,000 anonymized tickets from an enterprise-grade issue tracking system. Comparison reflects Standard Vector retrieval vs. Aspected retrieval pipelines using the same GPT-4o baseline. Skeptical? Request a live demo to verify the raw logs.

Introducing the Aspect Database

The Aspect Database is a retrieval database that indexes knowledge across multiple contextual dimensions (“aspects”), instead of relying only on embeddings.

We define Knowledge Objects, discrete units of enterprise knowledge indexed across six first-class, orthogonal dimensions. Each dimension is structured, filterable, and rankable at query time.

The Aspect Database structures knowledge across six dimensions:

1. Semantic Meaning

2. Intent

3. Process Position

4. Role Relevance

5. Authority & Validity

6. Constraints & Risk

 

Vector databases optimize for similarity.


The Aspect Database optimises for context-aware relevance.

Organisations that deploy Aspected already today

Aspected Customer Agfa

Customers

AGFA Service engineers use AI to gather, prepare, and orchestrate enterprise knowledge, delivering the right answer from the right source, directly within their ticketing system.

Customers
Aspected Partners adding value to our AI retrieval solution

Partners

Aspected works with different local and global AI consultancy and system integration partners. Such as Novadoc and Capgemini.

AI that works.

As aspected.

Aspected Logo - Icon Blue - Text Black

's-Gravelandseweg 178
1213 GB, Hilversum
The Netherlands
+31 35-6229545
https://aspected.com