Daily recap for June 25, 2026

Today we shipped 8 new step-by-step tutorials for small-business AI. Each one comes with a downloadable zip you can run in your own environment — no demos, no sales calls.

New tutorials

OpenAI Lead Intake with Intent Routing for SMBs

This lead intake agent classifies inbound chat messages by intent — pricing, demo, support — and routes qualified conversations to the right sales person. Full session history is preserved with compression, so returning leads pick up where they left off. Structured records push straight to your CRM.

Read the tutorial → Download the code (zip)

Built with @reaatech/confidence-router, @reaatech/confidence-router-classifiers, @reaatech/session-continuity, @reaatech/agent-handoff, on OpenAI · 102 tests, 97.09% coverage.

AWS Bedrock Contract Clause Extraction for SMB Legal

This pipeline extracts key clauses, dates, and parties from PDF contracts using AWS Textract for OCR and Bedrock for summarizing 12 standard clause types. It automatically repairs malformed JSON output and tracks per-contract cost against a daily budget, so small law firms can stop manually reviewing every agreement.

Read the tutorial → Download the code (zip)

Built with @reaatech/media-pipeline-mcp-doc-extraction, @reaatech/structured-repair-core, @reaatech/llm-cost-telemetry, on AWS Bedrock · 127 tests, 99.60% coverage.

Vertex AI Agent Mesh for SMB Multi-Channel Marketing

This mesh of specialized agents — content writer, brand reviewer, scheduler — uses Vertex AI to draft, review, and publish marketing content across channels. A confidence score gates each handoff: if the review scores too low, the content loops back for revision before it ever reaches LinkedIn. The full mesh state is observable in a live dashboard.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-mesh, @reaatech/agent-mesh-registry, @reaatech/agent-mesh-router, @reaatech/agent-mesh-confidence, @reaatech/agent-mesh-observability, on Vertex AI · 79 tests, 97.30% coverage.

Automated Review Response Agent for Independent Restaurants

This agent normalizes reviews from Yelp, Google, and TripAdvisor, runs them through a guardrail chain, and generates consistent on-brand replies. It stores memory for returning customers and records cost per response, so restaurant owners can maintain their online reputation without manual effort.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-mesh, @reaatech/guardrail-chain, @reaatech/llm-router-core, @reaatech/llm-cost-telemetry, @reaatech/mcp-server-core, @reaatech/agent-memory, on agnostic · 92 tests, 100.00% coverage.

Predictive Reservation Cancellation Backfill for Small Hospitality Groups

When a guest cancels or no-shows, this system predicts cancellation risk for upcoming reservations, activates the waitlist via SMS, and orchestrates AI voice calls to fill the table. It uses Twilio telephony, a voice agent, and caching to keep costs low.

Read the tutorial → Download the code (zip)

Built with @reaatech/voice-agent-core, @reaatech/voice-agent-telephony, @reaatech/agent-mesh, @reaatech/agent-handoff, @reaatech/mcp-server-tools, @reaatech/llm-cache, on agnostic · 109 tests, 93.17% coverage.

Mistral AI Lead Intake for ActiveCampaign SMB Lead Scoring

This lead intake pipeline qualifies and scores leads from web forms, chat, or email using Mistral AI, then pushes prioritized contacts to ActiveCampaign. It enforces a per-lead LLM spend budget and repairs any malformed output, so lean teams don't waste time on unqualified contacts.

Read the tutorial → Download the code (zip)

Built with @reaatech/confidence-router, @reaatech/agent-budget-engine, @reaatech/llm-cost-telemetry, @reaatech/structured-repair-core, on Mistral · 59 tests, 99.36% coverage.

OpenAI Voice Agent for Square Online Order Status Inquiries

This voice agent answers inbound calls via Twilio, transcribes speech, and looks up real-time Square order status. It handles ambiguous queries with confidence routing and maintains conversation state, so customers get quick answers without calling the owner.

Read the tutorial → Download the code (zip)

Built with @reaatech/voice-agent-core, @reaatech/voice-agent-stt, @reaatech/voice-agent-tts, @reaatech/voice-agent-telephony, @reaatech/session-continuity, @reaatech/confidence-router, @reaatech/llm-cost-telemetry, on OpenAI · 76 tests, 100.00% coverage.

Vertex AI Knowledge Agent for BigQuery SMB Data Queries

This RAG-powered knowledge agent syncs BigQuery data into a vector store and answers natural language questions via Vertex AI. It includes semantic caching, per-user budget enforcement, session continuity for multi-turn conversations, and confidence routing that escalates low-certainty answers. Non-technical staff can finally query their business data without writing SQL.

Read the tutorial → Download the code (zip)

Built with @reaatech/agent-memory, @reaatech/session-continuity, @reaatech/llm-cache, @reaatech/llm-cost-telemetry, @reaatech/agent-budget-engine, @reaatech/confidence-router, on Vertex AI · 115 tests, 95.89% coverage.

Browse all solutions →

Browse the full catalog at reaatech.com/products.