Daily recap for June 3, 2026

Today we shipped 15 new step-by-step tutorials for small-business AI, each a complete, downloadable recipe you can try yourself. They cover everything from generating a month of social media content from release notes to automatically redacting PII from LLM calls for HIPAA compliance.

New tutorials

Product Launch to Social Media Calendar

A five-person B2B startup can turn product release notes into a full 30-day social media calendar, with platform-specific copy for LinkedIn, Twitter, and Facebook. Instead of a marketing lead spending two days a month manually writing posts, the pipeline ingests Markdown notes and generates draft posts ready for scheduling. The whole thing runs on any OpenAI-compatible LLM.

Read the tutorial → · Download the code (zip)

Provider-agnostic, document-pipeline, Next.js, 85 tests, 98.55% coverage.

Support Ticket to Help Center Article Generator

An e-commerce company handling 200 support tickets a week can automatically turn closed Zendesk tickets into draft help center articles. The system clusters tickets by topic, extracts common questions and answers, and writes article drafts ready for review. It uses hybrid RAG across a Postgres + Qdrant stack, so the generated content stays grounded in real support data.

Read the tutorial → · Download the code (zip)

Provider-agnostic, knowledge-agent, Next.js, 173 tests, 99.34% coverage.

Agent Feedback Loop for Automated Fine-Tune Dataset Generation

A marketing agency that regularly corrects its AI-generated ad copy can collect those human edits into a clean fine-tuning dataset. The pipeline logs agent outputs, captures user accept/reject/edit decisions, and periodically exports a dataset for training a smaller, cheaper model that better matches the brand voice. Over time, it reduces reliance on expensive API calls.

Read the tutorial → · Download the code (zip)

Provider-agnostic, knowledge-agent, Next.js + NestJS, 113 tests, 98.18% coverage.

After-Hours Pet Emergency Triage Agent for Small Vet Practices

Small veterinary practices can automate after-hours phone triage with a voice agent that asks structured questions and assigns urgency levels. A speech-to-text pipeline feeds into an LLM that classifies the situation, and text-to-speech reads the recommendation back to the caller. The vet receives only the true emergencies.

Read the tutorial → · Download the code (zip)

Provider-agnostic, voice-agent, Next.js + Hono, 74 tests, 99.24% coverage.

Auto-repair call intake agent for service advisors

Independent auto-repair shops can stop losing bay time to estimate calls. This voice agent handles the initial “how much to fix X” conversation, identifies the repair type, provides a ballpark estimate, and only escalates complex requests to a human advisor. It wires together Twilio, Deepgram, ElevenLabs, and an LLM for repair assessment.

Read the tutorial → · Download the code (zip)

Provider-agnostic, voice-agent, Next.js + Fastify, 119 tests, 100.00% coverage.

Per-tenant LLM cost tracking and chargeback for vertical SaaS

B2B SaaS companies can finally bill tenants for their actual LLM usage. This system attributes every API call to the right tenant, computes real-time spend, and exports cost data to billing infrastructure. It uses OpenTelemetry and a custom cost pipeline that runs entirely in-memory until you plug in external exporters.

Read the tutorial → · Download the code (zip)

Provider-agnostic, cost-control, Next.js + Fastify, 74 tests, 100.00% coverage.

Anthropic Knowledge Agent for Shopify Product Q&A

Store owners on Shopify can deploy a conversational assistant that answers customer questions about products, inventory, and orders using live store data. It combines hybrid RAG over a synced product database with Anthropic's Claude, session continuity across visits, and long-term user preference memory. The agent knows what you have in stock and can reference past conversations.

Read the tutorial → · Download the code (zip)

Anthropic, knowledge-agent, Next.js, 111 tests, 99.64% coverage.

xAI Grok Observability for SMB AI Workflow Monitoring

If your SMB app uses Grok, you can now get a single-pane dashboard for token usage, latency, errors, and cost. This observability layer wraps every API call with OpenTelemetry spans and cost telemetry, then exports everything to Langfuse. No more guessing what Grok is costing you or when it fails.

Read the tutorial → · Download the code (zip)

xAI Grok, observability, Next.js, 59 tests, 96.64% coverage.

Automated Industry Newsletter Curator

A three-person consulting firm can replace its manual newsletter curation with an automated pipeline that fetches RSS feeds, scores articles by relevance, generates AI summaries, and compiles a draft ready for review. It uses a hybrid RAG stack to match articles to the firm's topic focus, cutting a five-hour weekly task to near zero.

Read the tutorial → · Download the code (zip)

Provider-agnostic, document-pipeline, Hono, 92 tests, 97.79% coverage.

xAI Grok Code Sandbox for Small Business Financial Modeling

Small business owners can generate financial models by describing them in plain English, then run the resulting code in an isolated sandbox with strict safety policies. The system uses xAI Grok to generate Python or JavaScript, executes it in an E2B sandbox, and exports results to a spreadsheet. Sensitive data stays protected by a tool-use firewall.

Read the tutorial → · Download the code (zip)

xAI Grok, code-execution, Express, 97 tests, 97.74% coverage.

Mistral AI Document Pipeline for NetSuite Invoice Automation

Finance teams that manually key invoice data from PDFs into NetSuite can automate extraction and posting. The pipeline ingests invoice PDFs, runs OCR with unpdf and tesseract.js, parses structured data with Mistral AI, repairs any malformed JSON, and posts draft vendor bills to NetSuite via its REST API. It also deduplicates invoices using agent memory.

Read the tutorial → · Download the code (zip)

Mistral, document-pipeline, Next.js, 99 tests, 100.00% coverage.

Configurable PII redaction at ingress and egress for HIPAA, GDPR, and PCI compliance

A health-tech startup handling protected health information can automatically strip PII from LLM calls and responses, with regulation-specific profiles for HIPAA, GDPR, or PCI. The dual-pipeline architecture redacts sensitive data before it reaches the model and scans responses before the user sees them, all with audit logging. No API keys required for the regex-based detection.

Read the tutorial → · Download the code (zip)

Provider-agnostic, security-guardrails, Hono + Fastify, 125 tests, 98.59% coverage.

Review Response Agent for Independent Restaurants

Restaurant owners stuck replying to Yelp, Google, and TripAdvisor reviews can automate consistent, on-brand responses. The agent analyzes the review text, checks against the restaurant's style guide, and drafts a reply that a human can approve or edit. It handles multi-turn conversations and caches frequent reply patterns to reduce latency and cost.

Read the tutorial → · Download the code (zip)

Provider-agnostic, knowledge-agent, Next.js + Fastify, 108 tests, 95.37% coverage.

Automated Blog Content Pipeline with SEO Scoring

A SaaS company that publishes two blog posts a week can go from topic brief to SEO-gated draft without an agency. The pipeline generates content, validates Markdown, lints for style, scores SEO factors like keyword density and readability, and produces a final draft that passes predefined quality gates. All behind a single API endpoint.

Read the tutorial → · Download the code (zip)

Provider-agnostic, document-pipeline, Next.js, 89 tests, 100.00% coverage.

Slack/email triage to priority inbox

A 20-person e-commerce support team can stop their first hour of manual message triage. This agent reads all inbound Slack DMs and Gmail emails, classifies each as P0 (urgent), P1 (high priority), or P2 (general), and drafts contextual replies for urgent items. It uses intent classification and confidence-gated routing to decide when to escalate to a human.

Read the tutorial → · Download the code (zip)

Provider-agnostic, lead-intake, Hono, 86 tests, 99.58% coverage.

Browse the full catalog at reaatech.com/products.