Daily recap for May 26, 2026

Today we shipped three new step-by-step tutorials for small-business AI: a Buildium knowledge agent, a cost control system for xAI Grok support bots, and an evaluation harness to keep your AI agents from breaking.

New tutorials

xAI Grok Knowledge Agent for Buildium Property Management Queries

Property managers and tenants can now get instant answers about leases, rent balances, and maintenance schedules without digging through Buildium’s portal. This tutorial walks you through connecting a Next.js API route to a RAG pipeline that indexes Buildium data and uses xAI Grok to respond conversationally.

Read the tutorial → Download the code (zip)

Built with xAI Grok, REAA hybrid-rag, agent-mesh, llm-router-core, and session-continuity; 116 tests, 99.3% coverage.

xAI Grok Cost Control for SMB Customer Support Agents

Small businesses running AI customer support agents often face unpredictable monthly bills. This tutorial shows how to add per-tenant daily spending caps, automatic fallback to a cheaper model when budgets are exhausted, and real-time cost telemetry—so your support stays online without overspending.

Read the tutorial → Download the code (zip)

Built with xAI Grok, REAA agent-budget-engine, llm-router-core, and helicone-async; 100 tests, 99.6% coverage.

Perplexity Agent Eval Harness for SMB AI Quality Assurance

If you’re deploying AI agents for chat or email, you need to know when an update breaks quality. This tutorial builds a CLI pipeline that uses Perplexity as a neutral judge to score your agent’s responses, gates prompt updates on evaluation results, and streams metrics to Langfuse—so you can ship confidently without expensive manual testing.

Read the tutorial → Download the code (zip)

Built with Perplexity, REAA perplexity-eval-glue, agent-eval-harness-suite, and prompt-version-control; 122 tests, 99.7% coverage.

Browse the full catalog at reaatech.com/products.