skills.rest
kernex-dev
kernex-devOFFICIAL
Explore all skills by this author →
WebsiteGitHub

ai-engineer

Design and implement RAG pipelines and LLM integrations for production AI systems.

1|Updated Mar 6, 2026
Search Tags:#embeddings#mlops#llm#rag#retrieval#vector-db#ai-engineer
One-click install
npx skills add https://github.com/kernex-dev/kernex-agent --skill ai-engineer-kernex-dev
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ai-engineer
Source: https://github.com/kernex-dev/kernex-agent/tree/main/builtins/ai-engineer
Command: npx skills add https://github.com/kernex-dev/kernex-agent --skill ai-engineer-kernex-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enable engineering teams to design, implement, and operate reliable AI features such as RAG pipelines, embedding search, LLM integrations, and production inference workflows that are maintainable, cost-aware, and privacy-safe.

Core Features & Use Cases

  • LLM Integration: Guidance for integrating OpenAI, Anthropic, Ollama, and Hugging Face with structured output, retries, and cost controls.
  • RAG & Embeddings: End-to-end patterns for chunking, embedding generation, vector indexing, retrieval, and re-ranking with citation-aware generation.
  • Vector Stores & MLOps: Practical advice for Qdrant, Pinecone, ChromaDB, and pgvector plus model versioning, monitoring, and automated retraining.
  • Production Hardening: Recommendations for rate limiting, fallback models, PII-safe logging, latency and cost metrics, and input/output validation.
  • Use Case: Build a document search and answer service that ingests PDFs, creates embeddings, indexes to a vector DB, and serves low-latency answers with source citations.

Quick Start

Create a retrieval-augmented generation pipeline that indexes your documentation with embeddings, stores vectors in Qdrant, and implements a re-ranking step before generation.

Frequently Asked Questions about ai-engineer

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

FAQPage Schema
How do I build a production-ready RAG pipeline with embeddings and a vector database?▼

To build a RAG pipeline, you implement document chunking, generate embeddings, index vectors in a vector database like Qdrant or Pinecone, and add a re-ranking step before LLM generation. This ensures accurate retrieval and citation-aware answers.

What is the best way to integrate LLM APIs like OpenAI and Anthropic with retry and cost controls?▼

Integrating LLM APIs involves connecting to OpenAI, Anthropic, or Hugging Face with structured output parsing, automated retries, and cost tracking. Implementing fallback models and rate limiting ensures reliable, cost-aware inference for backend applications.

Can I use pgvector or ChromaDB for embedding-based search in my AI application?▼

Yes, you can use pgvector or ChromaDB for embedding-based search. These vector stores integrate with retrieval pipelines to index and query embeddings, enabling low-latency document search and retrieval-augmented generation.

How do I implement PII-safe logging and monitoring for LLM inference workflows?▼

Implementing PII-safe logging and monitoring for LLM inference requires input/output validation, sensitive data redaction, and tracking latency and cost metrics. This protects user privacy while maintaining observability across MLOps workflows.

What are the key steps to set up MLOps workflows for model versioning and automated retraining?▼

Setting up MLOps workflows for model versioning and automated retraining involves tracking embedding model versions, monitoring retrieval and generation performance, and triggering automated retraining when data drift or latency degrades.

Does this approach support local LLM integrations using Ollama for privacy-sensitive applications?▼

Yes, local LLM integrations using Ollama are supported for privacy-sensitive applications. This allows you to run inference pipelines locally, reducing external API dependencies while maintaining structured output and rate limiting controls.

Related Skills

Explore More Software Engineering Skills→
InverterNetworkInverterNetwork

llama-cpp

Discovers GGUF models from Hugging Face Hub and configures llama.cpp inference servers.

Official
Advanced
Patasse97Patasse97

senior-ml-engineer

Deploy production ML systems with monitoring, drift detection, and cost controls.

Community
Advanced
patricio0312revpatricio0312rev
5

api-security-hardener

Apply rate limiting, input validation, authentication, and security headers to Node.js REST APIs.

Community
Advanced
OthmanAdiOthmanAdi
20

openui-forge-elixir

Integrate OpenUI generative UI into existing Elixir Phoenix projects.

Community
Intermediate