Catalog Catalogs
Catalogs
One dataset of 1,005 systems, as of 2026-10-31, viewed one audience at a time. Each catalog shows the rows, the columns and the decision tools that matter for its job, and nothing is left out: the Atlas is every system and every column in one table.
By layer of the stack
- Agent Memory 311 systems · 13 sections· 41 need attention Shipping memory layers, memory features and memory products for AI agents. For builders choosing how their agent remembers
- Agent Frameworks & Harnesses 238 systems · 10 sections· 43 need attention Frameworks, harnesses, IDEs and runtimes that agents are built and run in. For builders choosing an agent stack
- Retrieval & Data Infrastructure 58 systems · 5 sections· 9 need attention Vector stores, knowledge graphs, embeddings and search that memory is built on. For builders picking storage and retrieval substrate
- Models, Serving & Ops 134 systems · 6 sections· 25 need attention Foundation models, inference, training, evaluation and observability around agents. For teams operating agents in production
- Memory Research & Benchmarks 264 systems · 4 sections· 6 need attention Papers, benchmarks and ideas behind agent memory, before they ship as products. For researchers and teams evaluating memory claims
By what you are building
These cut across the layers. A system can appear in more than one.
- Coding Agents 97 systems · 5 sections· 25 need attention IDEs, coding harnesses, the memory they keep and the sandboxes they run in. For developers choosing or building a coding agent
- Security, Governance & Identity 42 systems · 4 sections· 9 need attention Guardrails, memory governance, agent identity and MCP gateways. For security reviewers and platform teams putting agents into production
- Evals & Observability 72 systems · 3 sections· 11 need attention Evaluation platforms, tracing, memory monitoring and the benchmarks behind them. For teams deciding how to test and watch agents in production
- Voice, Computer-use & Embodied Agents 52 systems · 4 sections· 11 need attention Voice agent platforms, wearable memory, computer-use agents and robotics stacks. For builders of agents that speak, see a screen or act in the world
- Assistant & Personal Memory 47 systems · 3 sections· 3 need attention How ChatGPT, Claude, Gemini and personal-knowledge tools remember you. For product teams and users comparing the memory inside assistants
- Vertical & Domain Agents 150 systems · 2 sections· 22 need attention Agents and memory built for one industry: legal, health, support, sales, science, gaming. For buyers and builders in a specific domain
The Atlas
Every system in one table, with every column behind the "All columns" toggle, every filter, export and comparison.
Add a catalog of your own by editing taxonomy/silos.json: a list of sections, the columns to show first, and the tools that apply.