framework-first · honest providers

Watch raw text become structured memory.

A framework-first RAG Memory engine. Real LangGraph orchestration; LlamaIndex, Mem0, OpenAI, and Langfuse promote per key, else fall back to honest stubs — surfaced in every run.

Open Memory Lab Compare configs

LangGraphLlamaIndex.TSMem0OpenAIRagas-shaped evalLangfusepgvector

open

Visual Memory Lab

Raw note → multi-level long-term memory. Normalize, classify, extract entities, split blocks, embed, store, link graph, consolidate.

Enter

open

Gold Memory Lab

Case study: 100 years of gold history as a typed memory graph. Ask, compare RAG vs memory, run the Golden Eval. Not investment advice.

Enter

open

Side-by-side Compare

Same query across 2–4 pipeline configs. Diff quality × cost × latency, winners per axis.

Enter

open

Framework Playground

Run any query through the LangGraph workflow. Inspect the full ExplainableRun trace.

Enter

open

Golden Eval

Score retrieval + answers against a golden set. Deterministic or LLM-as-judge.

Enter