AI Universe · Place the new thing
Place the new thing
A new name appears. Before reading its docs, ask four questions. Most of the time you discover you already understand it: it is a known concept, done a new way.
32 names, already placed
| Name | Family | Bucket | Kind | You already know it as |
|---|---|---|---|---|
| Ollama | Compute & infrastructure | Serving engine | Wrapper | an easy way to run open models locally (over llama.cpp) |
| vLLM | Compute & infrastructure | Serving engine | Implementation | software that serves a model fast |
| Pinecone | Data | Knowledge and retrieval | Implementation | a hosted vector database |
| RAG | Data | Knowledge and retrieval | Pattern | not a product: the pattern of retrieving passages into the context |
| pgvector | Data | Knowledge and retrieval | Implementation | a vector store inside PostgreSQL |
| Databricks | Data | Lakehouse layers | Bundle / platform | a platform spanning data, models, serving, agents and governance |
| LoRA | Model | Fine-tuning and alignment | Implementation | fine-tuning by training a small add-on, not the whole model |
| Llama 3 / gpt-oss | Model | The models | Model | open-weight models you can host yourself |
| Azure API Management | Model access | Gateway | Implementation | a general API gateway with AI policies |
| LiteLLM | Model access | Gateway | Wrapper | one OpenAI-shaped door over many providers' APIs |
| Groq | Model access | Inference API | Bundle / platform | custom chips plus an inference API, sold as a service |
| Microsoft Foundry | Model access | Inference API | Bundle / platform | Azure's platform of models, agents, evaluation and safety |
| OpenRouter | Model access | Routing and model choice | Bundle / platform | one key, many models, with routing |
| AWS Bedrock AgentCore | The AI application (agent harness) | Control loop | Bundle / platform | AWS's set of harness parts: runtime, memory, identity, gateway, observability |
| Claude Agent SDK | The AI application (agent harness) | Control loop | Bundle / platform | a loop, tools, context and permissions, packaged by one vendor |
| Google ADK | The AI application (agent harness) | Control loop | Implementation | Google's agent loop library |
| LangGraph | The AI application (agent harness) | Control loop | Implementation | a control loop plus state, written as a graph |
| Microsoft Foundry Agent Service | The AI application (agent harness) | Control loop | Bundle / platform | a hosted harness: loop, tools, memory, tracing on Azure |
| Pydantic AI | The AI application (agent harness) | Control loop | Implementation | a typed loop and a typed model layer |
| DBOS | The AI application (agent harness) | Durable work and people | Implementation | durable workflows that wait for a person |
| Temporal | The AI application (agent harness) | Durable work and people | Implementation | the same concept as DBOS, with its own server |
| Azure AI Content Safety | The AI application (agent harness) | Guardrails and policy | Implementation | harm and jailbreak checks, as an Azure service |
| Bedrock Guardrails | The AI application (agent harness) | Guardrails and policy | Implementation | input and output checks, as an AWS service |
| A2A | The AI application (agent harness) | Tools and protocols | Protocol / standard | the same idea as MCP, but agent to agent |
| MCP | The AI application (agent harness) | Tools and protocols | Protocol / standard | an agreed plug shape between an agent and its tools |
| Claude Code | Experience | Coding agents | Bundle / platform | a coding agent: an experience built on a full harness |
| Databricks Genie | Experience | Conversational analytics | Implementation | ask your data in words, over a semantic layer |
| DeepEval | Assurance & operations | Evaluation | Implementation | a library of graders and test cases |
| LangSmith | Assurance & operations | Evaluation | Bundle / platform | tracing, datasets and evaluation in one hosted product |
| Langfuse | Assurance & operations | Observability | Implementation | traces of every step, plus scores (it also covers evaluation) |
| OpenTelemetry | Assurance & operations | Observability | Protocol / standard | the standard way any code reports spans and metrics |
| PyRIT | Assurance & operations | Security and red-teaming | Implementation | an automated attacker for your agent |
Some names sit in several buckets: a bundle like Microsoft Foundry or Databricks spans many families. Place it by the job you are using it for.