skills.rest
jsdfhasuh
jsdfhasuhCOMMUNITY
Explore all skills by this author →
GitHub

rag-architect

Define production RAG architectures with hybrid retrieval, reranking, and evaluation metrics.

Updated Aug 23, 2026
Search Tags:#semantic search#rag#chunking#reranking#vector databases#hybrid retrieval#retrieval evaluation
One-click install
npx skills add https://github.com/jsdfhasuh/emosonic --skill rag-architect-jsdfhasuh
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-architect
Source: https://github.com/jsdfhasuh/emosonic/tree/main/.agents/skills/rag-architect
Command: npx skills add https://github.com/jsdfhasuh/emosonic --skill rag-architect-jsdfhasuh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

RAG systems often retrieve irrelevant context or generate answers that are not properly grounded in evidence, causing poor accuracy, high latency, and rising cost.

Core Features & Use Cases

  • End-to-end RAG system architecture: Covers ingestion, vector store design, retrieval pipelines, reranking, and evaluation loops for production deployments.
  • Retrieval quality engineering: Defines hybrid search, chunking strategies, reranking requirements, metadata filtering, and retrieval metrics (precision@k, recall@k, MRR, NDCG).
  • Operational scalability & iteration: Specifies idempotent ingestion, embedding versioning/migration, and continuous monitoring of retrieval latency and quality.
  • Use Case: Build a knowledge-grounded assistant for an enterprise support team by designing chunking and retrieval that consistently return the most relevant policy or troubleshooting documents before the LLM answers.

Quick Start

Use this skill to produce a full RAG architecture plan—including ingestion design, vector database selection, chunking strategy, retrieval flow, and an evaluation/monitoring checklist—for your target document types and scale.

Frequently Asked Questions about rag-architect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I design a production-grade RAG architecture that prevents irrelevant context retrieval?▼

Production-grade RAG architecture requires hybrid vector and keyword retrieval, metadata enrichment and filtering, and reranking chunks before LLM context augmentation to maintain high relevance and grounded outputs.

What chunking strategy should I use for semantic search in a vector database?▼

Effective chunking strategies for semantic search split documents into meaningful segments, which are then enriched with metadata and indexed in a vector database. The optimal chunk size depends on your specific document types and target query scope.

How do I evaluate retrieval quality and monitor RAG system performance?▼

Evaluate retrieval quality by tracking metrics like precision@k, recall@k, MRR, and NDCG. Monitor RAG system performance by continuously measuring retrieval latency and answer quality over time to catch degradation.

How to handle idempotent ingestion and embedding version migration in a vector database?▼

Handle idempotent ingestion by implementing deduplication during document indexing to prevent duplicate vectors. Plan embedding version migration by mapping and re-indexing documents when updating embedding models to avoid retrieval mismatch.

Does my retrieval-augmented generation pipeline need reranking before the LLM answers?▼

Reranking before the LLM answers is required to filter and reorder retrieved context chunks by relevance. This step prevents the LLM from processing low-priority or irrelevant context, improving accuracy and reducing latency.

What is the best way to build a knowledge-grounded assistant for enterprise support?▼

The best way to build a knowledge-grounded assistant is implementing an end-to-end RAG pipeline with hybrid search, metadata filtering, and continuous retrieval evaluation. This consistently returns relevant troubleshooting documents before LLM generation.

Related Skills

Explore More Software Engineering Skills→
neuron-coreneuron-core
2.0k

neuron-rag-specialist

Integrate vector stores, embeddings, and document loaders into Neuron AI agents.

Official
Advanced
anukkrit149anukkrit149

rag-implementation

Build Retrieval-Augmented Generation systems grounding LLM answers in vector databases.

Community
Advanced
patricio0312revpatricio0312rev
5

embedding-pipeline-builder

Automate document embedding pipelines for semantic search.

Community
Advanced
camoneartcamoneart
5

rag-implementation

Implement RAG systems with vector databases and retrieval strategies.

Community
Advanced