Topic
AI Agent Frameworks & MCP
LangChain, CrewAI, LlamaIndex, MCP: the orchestration layer wiring models to tools, memory, and data.
- Verified Facts
- 43
- Companies Tracked
- 4
- Business Signals (90d)
- 6
- Avg Fit Score
- —
Modules
Related topics
Other topics on the same stack layer (L7).
L7
Vector Databases
Pinecone vs Weaviate vs Qdrant vs Chroma: the retrieval layer every RAG and agent stack is built on.
L7
AI Eval & Observability
LangSmith, Arize, Langfuse, Braintrust: shipping agents to production needs eval + tracing.
L7
AI Gateways & LLM Routing
OpenRouter, LiteLLM, Portkey: the FinOps + routing layer controlling inference spend.
L7
Agent Memory
Stateful, temporal memory for AI agents — a category now distinct from vector-DB retrieval.
L7
Agent Runtime & Sandbox
Secure, ephemeral compute for AI agents to run code, tools, and browsers.
L7
Agent Security & Governance
Securing and governing AI agents — identity, guardrails, and red-teaming as agents gain autonomy.
Explore deeper
Engine-synthesised analysis for this topic, plus the ranked supplier leaderboard for its stack layer.
Tools
Subscribe to weekly updates on AI Agent Frameworks & MCP
One email Friday morning. Engineers only.
Subscribe to aiinframap Weekly →