jamwithai/production-agentic-rag-course

Features

• Agentic RAG Development - Provides a blueprint for building intelligent, self-correcting retrieval systems using agentic orchestration.
• Hybrid Search Retrievers - Combines semantic vector embeddings with keyword search using reciprocal rank fusion for improved retrieval.
• Agentic Workflow Automation - Uses intelligent agents to coordinate tasks and adapt retrieval based on result quality.
• Data Ingestion Pipelines - Automatically processes, chunks, and vectorizes raw data from APIs and PDFs for use in LLMs.
• Agentic Retrieval Workflows - Implements state-based agentic workflows for intelligent query rewriting and retrieved document grading.
• Ingestion Pipelines - Provides automated pipelines for extracting text from PDFs and external APIs to populate vector stores.
• RAG Pipelines - Implements workflows that augment model outputs by retrieving and integrating relevant external data from document sources.
• RAG Data Pipelines - Provides a practical blueprint for constructing data pipelines that process PDFs and store embeddings for semantic search.
• State-Based Workflow Engines - Manages complex agent execution flows using state graphs to coordinate query rewriting and document grading.
• Automated Document Ingestion - Automates the retrieval of external papers and transformation of PDFs into structured text for processing.
• Container Orchestration & Deployment - Provides configurations for managing application deployment within isolated containerized environments for databases and API services.
• Application Containerization - Packages software and dependencies into portable container images to ensure consistent execution across environments.
• RAG Development Courses - Offers a structured learning curriculum for building production-ready retrieval-augmented generation systems.
• LLM Architecture Guides - Provides a technical reference for implementing hybrid search, guardrails, and request tracing in LLM workflows.
• LLM Input Guardrails - Implements security layers that filter prompt injections and validate model inputs to prevent hallucinations.
• AI Guardrails - Implements validation layers that inspect inputs and detect domain boundaries to prevent hallucinations.
• LLM Response Streaming - Configures incremental delivery of language model outputs to reduce perceived latency for users.
• OCR Document Parsers - Uses optical character recognition to extract structured text and metadata from PDF files for vector indexing.
• PDF Text Extractors - Parses PDF files to extract structured text and metadata for use in RAG pipelines.
• Query Result Caching - Implements temporary storage of language model request results to reduce repeated computation and latency.
• Search Result Fusion Algorithms - Combines semantic vector embeddings with keyword search results using rank fusion to improve retrieval precision.
• AI Service Orchestration - Provides a framework for orchestrating containerized AI services, including databases, search engines, and LLM APIs.
• Containerized RAG Services - Employs deployment patterns for running RAG engines and their dependencies within containerized environments.
• Production Deployment Guides - Provides technical guidance for configuring high-availability production environments using container orchestration for RAG services.
• Server-Sent Events - Pushes real-time model outputs from the server to the client using standard event streams.
• Performance and Reliability - Implements retry logic, rate limiting, and local caching to ensure stable data ingestion pipelines.
• RAG Pipeline Optimizers - Implements request tracing and caching strategies to optimize retrieval latency and reduce operational costs.
• Request Tracing - Integrates observability tools to monitor end-to-end request paths and analyze system latency.
• Distributed Request Tracing - Captures timing and processing data across backend operations to analyze latency and operational costs.

Star history