Adaptive Agentic Rag With Dynamic Strategy Selection Based On Query Characteristics

via “retrieval-augmented generation (rag) pattern library with multiple retrieval strategies”

100+ AI Agent & RAG apps you can actually run — clone, customize, ship.

Unique: Provides 8+ distinct RAG patterns (basic, corrective, hybrid, database routing, agentic, autonomous, reasoning-enhanced) with working implementations for each, allowing developers to compare trade-offs between retrieval quality and latency. Most RAG tutorials show only basic vector search; this library treats RAG as a design space with multiple valid solutions.

vs others: More comprehensive RAG pattern coverage than LangChain's built-in RAG examples; more practical than academic RAG papers with runnable code for each pattern

via “adaptive-retrieval-with-query-routing”

This repository showcases various advanced techniques for Retrieval-Augmented Generation (RAG) systems. Each technique has a detailed notebook tutorial.

Unique: Implements query-aware routing that dynamically selects retrieval strategies based on query characteristics, allowing different query types to use optimized methods rather than forcing all queries through a single pipeline — an adaptive approach that improves both efficiency and quality

vs others: More efficient than applying all retrieval strategies to every query (fusion) because it selects the most appropriate strategy, and more effective than single-strategy systems because it adapts to query type

via “retrieval with multiple search strategies and vector database backends”

AutoRAG: An Open-Source Framework for Retrieval-Augmented Generation (RAG) Evaluation & Optimization with AutoML-Style Automation

Unique: Implements retrieval as a pluggable node type with multiple competing module implementations (BM25, semantic, hybrid, dense passage retrieval). Enables empirical evaluation of retrieval strategies and their impact on downstream answer quality without code changes.

vs others: More flexible than single-strategy retrieval because multiple strategies can be tested; more transparent than black-box retrieval because retrieved passages and scores are visible; enables strategy-selection based on empirical performance rather than assumptions.

via “multi-strategy query execution with global, local, and drift search”

A modular graph-based Retrieval-Augmented Generation (RAG) system

Unique: Implements three distinct search strategies (Global, Local, DRIFT) that operate at different abstraction levels of the knowledge graph, enabling adaptive retrieval based on query characteristics. DRIFT Search combines strategies with in-context fusion, allowing the LLM to reason over both community-level summaries and entity-level details in a single response.

vs others: More sophisticated than single-strategy RAG systems (e.g., basic vector similarity search), offering both breadth (global) and depth (local) reasoning. DRIFT Search's adaptive combination of strategies outperforms fixed-strategy approaches on diverse query types.

via “retrieval augmented generation system design and implementation”

A one stop repository for generative AI research updates, interview resources, notebooks and much more!

Unique: Organizes RAG design around explicit decision points (retriever type, embedding model, vector database, ranking strategy) with research-backed guidance on trade-offs. Includes dedicated section on agentic RAG patterns for knowledge-grounded agent decision making.

vs others: More comprehensive than framework-specific RAG documentation; provides cross-framework architectural patterns and research-backed design guidance, whereas most RAG resources focus on implementation in a specific framework.

via “agentic-rag-pattern-with-context-engineering”

12 Lessons to Get Started Building AI Agents

Unique: Frames RAG as an agentic decision (agents decide when to retrieve) rather than a static pipeline, and explicitly teaches context engineering techniques like chat summarization and scratchpad management to handle token constraints — most RAG tutorials treat retrieval as a fixed preprocessing step.

vs others: Covers the full context lifecycle (types, management, summarization) rather than just retrieval mechanics, making it more applicable to long-running agent conversations where context budgets are critical.

via “multi-strategy rag agent selection with automatic strategy routing”

Open Source Deep Research Alternative to Reason and Search on Private Data. Written in Python.

Unique: Implements three distinct RAG agent classes (NaiveRAG, ChainOfRAG, DeepSearch) with pluggable selection via configuration, enabling strategy swapping without code changes. DeepSearch agent specifically combines parallel search with LLM-based reranking and reflection loops — a pattern optimized for reasoning models like DeepSeek-R1 and Grok-3.

vs others: Offers more granular control over reasoning strategies than monolithic RAG systems; DeepSearch agent is specifically architected for reasoning models, whereas most RAG frameworks treat all LLMs equivalently

via “adaptive rag with query routing and dynamic context selection”

Ready-to-run cloud templates for RAG, AI pipelines, and enterprise search with live data. 🐳Docker-friendly.⚡Always in sync with Sharepoint, Google Drive, S3, Kafka, PostgreSQL, real-time data APIs, and more.

Unique: Implements query routing as a first-class pipeline component that dynamically selects retrieval strategies based on query classification, enabling cost and latency optimization without sacrificing answer quality. Supports both rule-based routing (fast, deterministic) and LLM-based routing (flexible, learned).

vs others: More sophisticated than basic RAG for high-volume systems; avoids the overhead of always retrieving context. Pathway's dataflow engine enables efficient routing without external orchestration frameworks.

via “tool-augmented-retrieval-with-query-expansion”

Agentic RAG is a different beast entirely.

Unique: Treats retrieval as a tool-calling problem where the agent selects and orchestrates multiple search strategies (semantic, keyword, graph, API) rather than relying on a single vector search backend, enabling richer query understanding

vs others: Outperforms single-backend RAG on diverse data types because it can route queries to appropriate tools (keyword search for exact matches, semantic search for conceptual similarity, APIs for real-time data) rather than forcing all queries through one retrieval method

Agentic-RAG explores advanced Retrieval-Augmented Generation systems enhanced with AI LLM agents.

Unique: Implements adaptive strategy selection where agents analyze query characteristics to determine optimal processing approach, rather than using uniform strategies for all queries, enabling efficient resource utilization by matching complexity to requirements.

vs others: More efficient than fixed-strategy systems by adapting to query characteristics, and more intelligent than simple routing by using query analysis to select strategies that balance multiple optimization objectives.

via “dynamic model selection based on context”

MCP server: mcp-server-test

Unique: Employs decision trees for real-time model selection based on context, enhancing relevance over static approaches.

vs others: More adaptive than static model routing systems, providing tailored responses based on user context.

via “dynamic model switching”

MCP server: mbit-test

Unique: Incorporates a decision-making layer that evaluates requests to select the most suitable model dynamically.

vs others: More efficient than static model setups, as it adapts to the specific needs of each request in real-time.

via “adaptive rag with query-dependent retrieval strategy selection”

Open-source Python library to build real-time LLM-enabled data pipeline.

Unique: Dynamically selects retrieval strategy based on query analysis, eliminating need for manual strategy selection. Integrates query analysis into the retrieval pipeline, enabling intelligent routing without separate preprocessing steps.

vs others: More effective than fixed retrieval strategies because it adapts to query characteristics; more efficient than trying all strategies because it selects the best one upfront.

via “dynamic model selection”

MCP server: facebook-gemini-agents

Unique: Employs a sophisticated decision-making algorithm that evaluates multiple models based on real-time performance metrics and user intent.

vs others: More adaptive than static model selection methods, providing tailored responses based on context.

via “dynamic model selection based on context”

MCP server: tcmb-mcp-server

Unique: Incorporates machine learning techniques for context analysis to improve model selection accuracy and efficiency.

vs others: More intelligent than static routing systems, as it adapts to user input and context for optimal model usage.

via “dynamic model selection based on user input”

MCP server: mcp-hackathon-africa

Unique: Incorporates real-time evaluation of user input to select models, providing a level of responsiveness that static systems lack.

vs others: More responsive than static model selection systems, which do not adapt to real-time user input.

via “dynamic model selection”

MCP server: mcp-server-251215

Unique: Incorporates a sophisticated criteria-based model selection process that adapts to user needs in real-time, unlike static model setups.

vs others: More efficient than fixed model setups, as it adapts to the specific requirements of each request.

via “dynamic model switching”

MCP server: dowhistle-mcp-server1

Unique: Employs a context-based decision-making algorithm that evaluates model performance in real-time, enhancing responsiveness.

vs others: More adaptive than static model deployment systems, as it can respond to varying user needs on-the-fly.

via “dynamic model selection based on user-defined criteria”

MCP server: shelf-mcp

Unique: Features a decision-making engine that evaluates user-defined criteria for model selection, which is a unique approach compared to static model invocation methods.

vs others: More adaptive than traditional MCPs that rely on pre-defined model calls without dynamic evaluation.

via “dynamic routing for model requests”

MCP server: smithery-mcp-server

Unique: Employs a sophisticated routing algorithm that adapts to user needs and model capabilities in real-time.

vs others: More efficient than static routing systems as it adapts to varying user needs and model performance.