How daethyra/build-ragai works
Build-RAGAI is an educational collection of LangChain, OpenAI, and Transformers examples rather than one integrated application. The principal workflow is a user exploring a notebook or runnable example to build an AI assistant or retrieval workflow; other branches cover document Q&A, speech transcription and task extraction, and model fine-tuning. Internal edges are included only where the sampled source shows a call. Several README capabilities are represented at their documented directory level because their implementations were not sampled.
Main entry points: application.py Pinecone workflow · main.py Pinecone app entry · run.py Microphone app runner · qa_local_docs.py Directory PDF Q&A
Components 28
LangChain learning
- DeepSeek agent
src/langchain/notebooks/DeepSeek.ipynb - LangChain notebooks
src/langchain/notebooks - Reusable snippets
src/langchain/codesnippets - Conversation memory
src/langchain/codesnippets/bufferwindow_memory.py - Chat model example
src/langchain/codesnippets/chatopenai.py - Multi-query retrieval
src/langchain/codesnippets/multi_queryvector_retrieval.py - Agent and tools tutorials
src/langchain/notebooks - Tracing tutorial
src/langchain/notebooks/langsmith_walkthrough.ipynb - Runnable and chain tutorial
src/langchain/notebooks/lcel_runnables_and_chains.ipynb
Retrieval applications
- Directory PDF Q&A
src/langchain/packages/rag-with-agents/directoryloader/qa_local_docs.py - Directory Q&A runner
src/langchain/packages/rag-with-agents/directoryloader/run_qa_local_docs.py - PyPDF Q&A
src/langchain/packages/rag-with-agents/pypdfdirectoryloader/query_local_docs.py - FAISS retriever
src/langchain/packages/rag-with-agents/faiss_retriever.py - Pinecone app entry
src/langchain/packages/vectorstore-rag/pinecone/main.py - Pinecone workflow
src/langchain/packages/vectorstore-rag/pinecone/application.py - Document loading
src/langchain/packages/vectorstore-rag/pinecone/documentservice.py - Vector retrieval
src/langchain/packages/vectorstore-rag/pinecone/vectorstoreservice.py - RAG notebook examples
src/langchain/notebooks - Vector-store tutorials
src/langchain/notebooks
Transformers applications
- Microphone app runner
src/transformers/packages/audiotranscription/mictranscription/run.py - Realtime transcription
src/transformers/packages/audiotranscription/mictranscription/transcribe_microphone.py - Speech task extraction
src/transformers/packages/audiotranscription/taskcreation/transcribe_tasks.py - Sequence-model fine-tuning
src/transformers/packages/trainwithaccelerate/fine_tune_sequence_classification_model.py - ASR notebook demo
src/transformers/notebooks/asr_pipelines.ipynb - Transformers snippets
src/transformers/codesnippets
OpenAI examples
- Speech generation
src/opai/codesnippets/tts.py - OpenAI RAG notebook
src/opai/notebooks/gen-qa-openai.ipynb
Other components
- Learner
Connections 22
- Learner → DeepSeek agent: explores agent
README.md - Learner → LangChain notebooks: studies examples
README.md - Learner → Reusable snippets: adapts components
README.md - Learner → Directory PDF Q&A: runs document Q&A
README.md - Learner → PyPDF Q&A: queries PDFs
README.md - Learner → Pinecone app entry: runs retrieval app
README.md - Learner → RAG notebook examples: learns RAG
README.md - Learner → Microphone app runner: starts transcription
README.md - Learner → Speech task extraction: extracts tasks
README.md - Learner → Sequence-model fine-tuning: fine-tunes model
README.md - Learner → ASR notebook demo: tries ASR demo
README.md - Learner → Speech generation: generates speech
README.md - Learner → OpenAI RAG notebook: learns OpenAI RAG
README.md - Learner → Vector-store tutorials: compares stores
README.md - Learner → Agent and tools tutorials: learns tool use
README.md - Learner → Tracing tutorial: learns tracing
README.md - Learner → Runnable and chain tutorial: learns chains
README.md - Directory Q&A runner → Directory PDF Q&A: invokes
src/langchain/packages/rag-with-agents/directoryloader/run_qa_local_docs.py - Pinecone app entry → Pinecone workflow: creates and runs
src/langchain/packages/vectorstore-rag/pinecone/main.py - Pinecone workflow → Document loading: loads documents
src/langchain/packages/vectorstore-rag/pinecone/application.py - Pinecone workflow → Vector retrieval: indexes and retrieves
src/langchain/packages/vectorstore-rag/pinecone/application.py - Microphone app runner → Realtime transcription: creates and captures
src/transformers/packages/audiotranscription/mictranscription/run.py