Today we shipped 8 new tutorials, including a Cohere knowledge agent that lets small business owners query their financial data in plain English—no SQL required.
New tutorials
Cohere knowledge agent for PostgreSQL-backed financial analytics
This tutorial builds a knowledge agent on Cohere that lets small business owners ask questions about transactions, expenses, and cash flow in plain English, getting answers from their PostgreSQL data without writing SQL. It handles multi-turn conversations, caches similar queries to cut costs, and asks for clarification when a question is ambiguous.
Read the tutorial → https://reaatech.com/solutions/cohere-knowledge-agent-for-postgresql-backed-financial-analytics
Download the code (zip) → artifact.zip
Built with @reaatech/agent-memory-core, @reaatech/llm-cache, @reaatech/confidence-router-core, @reaatech/structured-repair-core, on Cohere · 88 tests / 98.63% coverage.
Databricks voice agent for after-hours property maintenance intake
Property managers can set up a voice agent on Twilio that answers after-hours calls, uses Databricks to classify the issue, and creates AppFolio work orders automatically. Deepgram handles speech-to-text, and session continuity keeps context across calls.
Read the tutorial → https://reaatech.com/solutions/databricks-voice-agent-for-after-hours-property-maintenance-intake
Download the code (zip) → artifact.zip
Built with @reaatech/voice-agent-core, @reaatech/voice-agent-stt, @reaatech/voice-agent-tts, @reaatech/voice-agent-telephony, @reaatech/confidence-router-core, @reaatech/agent-handoff, @reaatech/session-continuity, on Databricks · 117 tests / 97.46% coverage.
OpenRouter budget guardrails for BigCommerce SMB customer support
Wrap your BigCommerce chatbot's LLM calls with a budget layer that caps daily spend, switches to cheaper OpenRouter models near limits, caches FAQs, and records every penny spent—so your support AI never surprises you with a bill.
Read the tutorial → https://reaatech.com/solutions/openrouter-budget-guardrails-for-bigcommerce-smb-customer-support
Download the code (zip) → artifact.zip
Built with @reaatech/agent-budget-engine, @reaatech/llm-cost-telemetry, @reaatech/llm-router-engine, @reaatech/llm-cache, on OpenRouter · 63 tests / 94.26% coverage.
1099-NEC contractor classification and W-9 collection agent
CPA firms can automate year-end 1099 reporting: this agent classifies contractors from payment data, sends W-9 requests via Gmail, parses responses, and generates 1099-NEC PDFs. It uses OpenAI and six REAA agent packages for handoff, memory, and cost control.
Read the tutorial → https://reaatech.com/solutions/openai-1099-collection-agent
Download the code (zip) → artifact.zip
Built with @reaatech/agent-handoff, @reaatech/agent-memory, @reaatech/agent-runbook, @reaatech/agent-eval-harness-suite, @reaatech/llm-cache, @reaatech/agent-budget-engine, on OpenAI · 130 tests / 97.58% coverage.
Automated receipt classifier for small CPA firms
Bookkeepers can upload receipts and this agent uses OCR and an LLM to extract vendor, amount, and GL code, validated by a guardrail chain. It tracks cost and evaluations so you can merge it into your existing workflow.
Read the tutorial → https://reaatech.com/solutions/agnostic-receipt-classifier-agent
Download the code (zip) → artifact.zip
Built with @reaatech/agents-markdown, @reaatech/agent-mesh, @reaatech/llm-router-core, @reaatech/agent-eval-harness-golden, @reaatech/guardrail-chain, @reaatech/agent-budget-spend-tracker, on Agnostic · 91 tests / 94.61% coverage.
xAI Grok lead intake for Mailchimp SMB email marketing
This tutorial sets up a lead intake system that uses xAI Grok to classify form submissions as hot, warm, or cold and automatically adds qualified leads to Mailchimp with the right tags, thanks to confidence routing and structured output repair.
Read the tutorial → https://reaatech.com/solutions/xai-grok-lead-intake-for-mailchimp-smb-email-marketing
Download the code (zip) → artifact.zip
Built with @reaatech/confidence-router, @reaatech/confidence-router-classifiers, @reaatech/structured-repair-core, @reaatech/llm-cost-telemetry, on xAI Grok · 79 tests / 100.00% coverage.
Anthropic document pipeline for Square SMB receipt extraction
This pipeline ingests receipt images, uses Anthropic Claude to extract structured data like vendor and total, enforces a daily cost budget, and pushes the results to Square—so you can stop entering receipts manually.
Read the tutorial → https://reaatech.com/solutions/anthropic-document-pipeline-for-square-smb-receipt-extraction
Download the code (zip) → artifact.zip
Built with @reaatech/confidence-router, @reaatech/structured-repair-core, @reaatech/llm-cost-telemetry, @reaatech/agent-budget-engine, on Anthropic · 137 tests / 97.42% coverage.
Anthropic multi-agent handoff for Clio SMB legal case management
This tutorial builds a three-agent system that triages client messages, drafts professional replies, and updates Clio case records—reducing non-billable hours for intake. A budget engine caps spending per agent.
Read the tutorial → https://reaatech.com/solutions/anthropic-multi-agent-handoff-for-clio-smb-legal-case-management
Download the code (zip) → artifact.zip
Built with @reaatech/agent-mesh, @reaatech/agent-mesh-router, @reaatech/agent-handoff, @reaatech/agent-handoff-routing, @reaatech/agent-budget-engine, on Anthropic · 153 tests / 99.20% coverage.
Browse all solutions → https://reaatech.com/solutions
Browse the full catalog at reaatech.com/products
- recap
Daily recap for June 21, 2026
We shipped 7 new tutorials: spend control for Databricks, lead intake for financial advisors, medical claim extraction, an approval-gated code sandbox, a voice agent for Buildium, Perplexity guardrails for QuickBooks, and a Stripe dispute response agent mesh.
- recap
Daily recap for June 20, 2026
Seven new download-and-run tutorials for small businesses: invoice automation for Magento, buyer-rep compliance, multi-agent support routing, and more.
- recap
Daily recap for June 19, 2026
Nine new tutorials landed today — from receipt automation and voice booking to reliability suites and on-prem evaluation, each with a step‑by‑step walkthrough and a ready‑to‑run zip.
Comments
Sign in with GitHub to comment and vote.