RAG Showcase - Six Approach Flow Phases

Parallel lane view of ingestion, retrieval, augmentation, generation, and scoring behavior

0. Shared setup and corpus preparation 0.1 load corpus 0.2 chunk docs 0.3 embed chunks 0.4 contextualize 0.5 upload LightRAG 0.6 register models 1. Entry 2. Retrieval plan 3. Evidence source 4. Augmentation 5. Generation 6. Output + metrics 7. Observed trait vanilla-rag hybrid-rag contextual-rag graph-rag agentic-rag n8n-adaptive-rag 1.1 /vanilla-ragOpenAI chat request 2.1 embed querynomic-embed-text 3.1 RagBasedense top-k=5 4.1 stuff contexttop chunks only 5.1 one LLM calllight_gen role 6.1 answersources + metrics baselinefast/simple 1.1 /hybrid-ragsame request shape 2.1 embed + textsemantic + BM25 3.1 RagBasehybrid retrieve k=20 4.1 TEI reranktop_n=5 5.1 one LLM callstuff reranked hits 6.1 answerrerank cost in latency keyword IDspredictable 1.1 /contextual-ragsame request shape 2.1 embed + textsemantic + BM25 3.1 RagContextualblurb-prefixed chunks 4.1 TEI reranktop_n=5 5.1 one LLM callcontext-aware hits 6.1 answerbest judged default robusttop score 1.1 /graph-ragdelegating endpoint 2.1 LightRAG querymode=hybrid 3.1 KG + vectorentities + relations 4.1 graph contextrerank disabled 5.1 LightRAG LLMmistral-small role 6.1 answerKG source marker sloweruneven 1.1 /agentic-ragresearch-agent prompt 2.1 ReAct loopMAX_STEPS=4 3.1 toolsvectors or graph 4.1 observationstool trace builds up 5.1 final answeror MAX_STEPS stop 6.1 answer + traceLLM call count varies flexiblestep-limited 1.1 /n8n-adaptive-ragthin webhook wrapper 2.1 classifysimple vs complex 3.1 routevanilla or agentic 4.1 selected pathHTTP Call Approach 5.1 workflow resultRespond to Webhook 6.1 answer + routeadaptive source marker operatorconfigurable Shared model layer: LiteLLM routes to the configured Atlas provider; generation uses roles.yaml; LightRAG uses Atlas role inputs; embeddings default to nomic-embed-text.

1. Three direct retrievers

  • vanilla, hybrid, and contextual share the same answer_from_context generation path.
  • The main difference is evidence selection: dense only, hybrid+rerank, or contextual chunks+rerank.

2. Graph and agent paths

  • graph-rag delegates query answering to LightRAG hybrid mode.
  • agentic-rag loops over vector search and LightRAG tools, then returns a trace and final answer.

3. Adaptive workflow

  • n8n-adaptive-rag is intentionally a workflow bridge, not a new retriever.
  • The operator-built n8n graph classifies the query and calls a selected approach.

Each approach lane is deployed as a mounted FastAPI route inside Atlas backend, registered as a LiteLLM model alias, then invoked by Open WebUI or compare/run_matrix.py through LiteLLM /v1/chat/completions.