ApertureData

This account is verified. Learn more

@ApertureData

Foundational Data Layer for AI: Combine scalable vector search with memory-optimized graph and multimodal data management

Mountain View, CA aperturedata.io/home Joined November 2018

30 Following 403 Followers 625 Posts

Latest posts 5

ApertureData @ApertureData 3d

ApertureDB array properties can be indexed. An array index can speed up membership checks and selected-element lookups.

["red", "in", "$colors"]
[{"at": ["$colors", 0]}, "==", "red"]

Docs:
docs.aperturedata.io/query_language…

ApertureData @ApertureData 6d

People type "Scarlet Johanson." A typo-tolerant full-text match on the name finds Scarlett Johansson as the entity, and the knowledge graph traversal can start there: films, co-stars, their other work.

Each query word gets its own max_edits. The search starts with exact matches, allows one more edit at a time up to the ceiling you set, and stops at the first step that returns results. Results are ranked with BM25.

The constraint, on a property with a fulltext index:
[{"fulltext": ["$name", "harbr lihgts", "all", 4]}, ">", 0]

Docs:
docs.aperturedata.io/query_language…

ApertureData @ApertureData Sep 14

An architecture walkthrough, an early look at an internal tool actually being built, and some of the most experienced builders in agentic memory in one room. San Francisco, October 13.

Vishakha Gupta-Cledat Verified account @vishakha041 Sep 14

Oct 13, San Francisco: an evening on agentic memory in practice.

I'll walk through the architecture behind memory and cognition in AI, grounded in a study of 20+ frameworks. Sonam Gupta (@Coffee_and_NLP ) shows an early look at a sales enablement tool she's building on Aperture Nexus (from @ApertureData ).

Hosted with @heyamylin @OutcastVC , with Connie Xu (Halo) bringing together builders in the space.

RSVP: luma.com/8tbyxmd8
View quoted post on X

ApertureData @ApertureData Sep 9

1/
An agent walks into a repeat problem like it's the first time, every time.
Aperture Nexus is our answer: open source memory for AI agents, built on ApertureDB (aperturedata.io). 🧵

ApertureData @ApertureData Sep 2

Memory is the dessert everyone in AI wants right now. Episode 1 of The Cognitive Layer makes the case for eating your vegetables first, the data foundation underneath it.

Himanshu Singh (Netflix) on why entity resolution is a systems problem, not a prompt problem, and what happens when you skip straight to "memory": agents don't just make mistakes, they hallucinate in sync.

Vishakha Gupta-Cledat Verified account @vishakha041 Sep 2

Asked my guest: coffee, cocktail, or dessert?

Him: "Dessert. Memory is the dessert everyone wants. But you have to eat your vegetables first, those are the data foundation. Skip them, and the dessert makes you sick."

Episode 1 of The Cognitive Layer, with Himanshu Singh (@netflix ), on building AI memory you can actually trust.

Link below 👇

View quoted post on X