From the Lab
Research
Latest · Jul 11, 2026
SAGE 1.10: A Technical Overview
An overview of the memory, context-management, observability, and concurrency changes introduced across the SAGE 1.6 through 1.10 releases, and the design reasoning behind each.
Read the post→SAGE Computer Use: Agents That Drive Real Software
A portable computer-use layer that browses, runs shell, controls a desktop, and verifies its own work, running entirely on local inference.
Workflows: Deterministic Pipelines for AI Agents
Agentic loops are powerful but unpredictable. When you know the steps in advance, you need a pipeline, not a prayer. Here's how we built deterministic workflow execution into SAGE.
Lattice: A Memory Engine That Actually Remembers
How We Built Persistent, Grounded Memory for AI Agents and Why Most Systems Get It Wrong
Why We Built SAGE: What Production AI Agents Actually Require
Most tutorials show you how to get an LLM to respond. The hard part is everything after. Here's what we learned building production agents, and why we built the engine that powers everything we ship.
Work with NeuroQuest Labs
Put this research
to work.
Everything we publish here ships in SAGE. Talk to us about building on it.