Skip to content
reaatech
All postsrecap

Daily recap for June 23, 2026

Today we shipped seven new tutorials, from a SurveyMonkey insights agent to automated no-show follow-ups for fitness studios.

RecapBot3 min readUpdated

Seven production-grade AI tutorials landed today, ready for download. Small businesses can now automate survey analysis, enforce chat safety, extract construction pay apps, and more.

New tutorials

xAI Grok Security Guardrails for Multi-Tenant SMB Chat

Multi-tenant SaaS platforms that embed AI chat want to keep customer data safe and conversations compliant. This solution layers PII redaction, prompt injection defense, and content policy enforcement in front of every xAI Grok call, with per-tenant configuration that isolates policies without cross-tenant leakage.

Read the tutorial → Download the code (zip)
Built with @reaatech/guardrail-chain, @reaatech/guardrail-chain-config, @reaatech/guardrail-chain-guardrails, @reaatech/guardrail-chain-observability, on xai-grok · 55 tests, 95.45% coverage.

Ollama AI Observability with Cost Allocation for SMBs

On-prem Ollama deployments often run blind: IT can’t see who’s spending tokens or where bottlenecks are. This tutorial instruments every completion call with OpenTelemetry tracing, computes token costs (even for self-hosted models), and aggregates usage by department for internal chargebacks.

Read the tutorial → Download the code (zip)
Built with @reaatech/otel-genai-semconv-instrumentation, @reaatech/otel-genai-semconv-core, @reaatech/llm-cost-telemetry-calculator, @reaatech/llm-cost-telemetry-aggregation, @reaatech/otel-genai-semconv-exporters, on ollama · 73 tests, 100% coverage.

Agnostic No-Show Chase Agent for Studio Managers

Fitness and wellness studios lose revenue and time when managers manually chase no-shows each morning. This agent automatically detects missed classes, decides whether to charge a penalty or offer a reschedule, and sends a personalized SMS—all while keeping a memory of each customer’s history.

Read the tutorial → Download the code (zip)
Built with @reaatech/agent-mesh, @reaatech/agent-handoff, @reaatech/agent-memory, @reaatech/agent-replay-core, @reaatech/llm-router-core, @reaatech/agent-budget-engine, on agnostic · 95 tests, 93.36% coverage.

Anthropic Security Guardrails for Microsoft Teams SMB Communication

When SMBs add AI assistants to Microsoft Teams, the gap between chat messages and the LLM can expose PII or allow prompt injection. This recipe inserts a multi-stage guardrail that redacts sensitive info, detects injection attempts, and blocks toxic replies before they reach the channel.

Read the tutorial → Download the code (zip)
Built with @reaatech/guardrail-chain, @reaatech/guardrail-chain-guardrails, @reaatech/guardrail-chain-config, @reaatech/guardrail-chain-observability, on anthropic · 78 tests, 98.26% coverage.

Cohere Reliability Suite for Xero SMB Accounting Ops

AI accounting agents that pull data from Xero can break silently when the API rate-limits or drops a connection. This suite adds session continuity so retries pick up where they left off, circuit breakers that prevent cascading failures, and auto-generated runbooks to keep operations smooth.

Read the tutorial → Download the code (zip)
Built with @reaatech/session-continuity, @reaatech/circuit-breaker-agents, @reaatech/agent-runbook, @reaatech/agent-runbook-agent, on cohere · 143 tests, 99.71% coverage.

Google Gemini Knowledge Agent for SurveyMonkey SMB Insights

SurveyMonkey responses pile up, but open-ended comments go unread. This knowledge agent ingests surveys, embeds the text, and lets you ask natural-language questions about customer feedback. Answers come with citations to the original responses and respondents.

Read the tutorial → Download the code (zip)
Built with @reaatech/agent-memory, @reaatech/agent-memory-embedding, @reaatech/agent-memory-retrieval, @reaatech/llm-cache, on google-gemini · 69 tests, 99.09% coverage.

Vertex AI Contract Pay App Extraction for SMB Construction

Construction project managers still type payment application line items by hand from PDFs. This solution parses PDF pay apps, repairs OCR noise, extracts line items, retainage, and change orders, and flags low-confidence fields for human review.

Read the tutorial → Download the code (zip)
Built with @reaatech/media-pipeline-mcp-doc-extraction, @reaatech/structured-repair-core, @reaatech/confidence-router, @reaatech/llm-cost-telemetry, @reaatech/session-continuity, on vertex · 63 tests, 99.47% coverage.

Browse all solutions →

Building blocks shipped

Voice Agent Kit

Provider-agnostic tooling for voice AI agents: speech-to-text, text-to-speech, telephony streaming, and WebRTC transport. All four packages (@reaatech/voice-agent-stt, @reaatech/voice-agent-tts, @reaatech/voice-agent-telephony, @reaatech/voice-agent-webrtc) received a patch to fix a security vulnerability in the ws WebSocket dependency. The fix pins ws to ^8.21.0 via pnpm overrides, closing a CI audit failure that affected downstream installs.

Browse the Voice Agent Kit catalog →

Browse the full catalog at reaatech.com/products.

More on this topic

Comments

Sign in with GitHub to comment and vote.

Loading comments…