Daily recap for May 13, 2026

Today we shipped six new step-by-step tutorials for small-business AI, each a complete working system you can download and try. They cover field service voice dispatch, safe code execution for financial modeling, AI spend tracking, automated invoice processing, Vertex AI budget guardrails, and AWS Bedrock incident runbooks.

New tutorials

Vertex AI Voice Agent for Field Service Dispatch

Answer after-hours field service calls automatically. This voice agent uses Vertex AI to classify requests—emergency repair, maintenance, or estimate—and routes them to the right on-call technician. It retains customer history, monitors spending per call, and runs health checks.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-handoff-routing, @reaatech/agent-mesh-session, @reaatech/agent-memory, @reaatech/agent-budget-engine, @reaatech/agent-runbook-health-checks, on Vertex AI · 86 tests / 94.6% coverage.

OpenAI Code Sandbox for SMB Financial Modeling

Run AI-generated Python code safely for financial analysis. This sandbox isolates execution in an E2B container, so customer-submitted code never touches your production environment. An intelligent router decides when to call OpenAI for code generation or execute directly, with gateway-enforced access controls and health monitoring.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-handoff-routing, @reaatech/agent-mesh-gateway, @reaatech/agent-runbook-health-checks, @reaatech/agent-chaos-core, on OpenAI · 94 tests / 99.4% coverage.

Anthropic AI Spend Tracker for SMB Agent Workflows

Track Anthropic API spend per agent with real-time dashboards and automatic budget enforcement. This system captures token usage, enforces per-agent caps, and downgrades models when thresholds are hit. A background sync loop pools batch usage data to keep cost records up to date.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-budget-spend-tracker, @reaatech/agent-budget-engine, @reaatech/agent-runbook-observability, @reaatech/agent-eval-harness-observability, @reaatech/agent-budget-cli, on Anthropic · 123 tests / 96.4% coverage.

Anthropic Document Pipeline for SMB Accounting

Automate invoice processing with high accuracy and error detection. Upload PDFs or images and a Redis-backed queue extracts structured data via Claude, validates against a schema, retries failures, and logs per-invoice costs. Built for accounting teams that spend hours rekeying data.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-mesh-gateway, @reaatech/agent-mesh-router, @reaatech/agent-handoff, @reaatech/agent-eval-harness-gate, @reaatech/agent-budget-engine, @reaatech/agent-runbook-agent, @reaatech/agent-memory, on Anthropic · 109 tests / 99.3% coverage.

Vertex AI Budget Guardrails for Multi-Agent Systems

Keep Vertex AI spend predictable with per-agent budget limits and automatic model routing. A middleware enforces budget rules on every LLM request, attaches cost headers, and blocks calls that exceed limits. When cheaper models are available, the router switches automatically using Vertex-specific rate mapping.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-budget-engine, @reaatech/agent-budget-llm-router-plugin, @reaatech/agent-budget-spend-tracker, @reaatech/agent-budget-middleware, @reaatech/agent-budget-pricing, @reaatech/agent-eval-harness-cost, on Vertex AI · 40 tests / 95.2% coverage.

AWS Bedrock AI Incident Runbook for DevOps-light SMBs

Automate incident response for AI agents so lean teams can sleep. This runbook system monitors services, detects failures, and uses Bedrock to generate plain-language triage reports. It classifies incidents against known AI failure modes, triggers automatic rollbacks for critical issues, and sends Slack notifications — all from a dashboard your team can customize in natural language.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-runbook, @reaatech/agent-runbook-agent, @reaatech/agent-runbook-alerts, @reaatech/agent-runbook-failure-modes, @reaatech/agent-runbook-observability, @reaatech/agent-runbook-rollback, @reaatech/agent-runbook-service-map, on AWS Bedrock · 47 tests / 100% coverage.

Browse all solutions →