Building a Production WhatsApp AI Agent: Architecture That Actually Works

A developer has detailed the architecture behind SARA, an open-source WhatsApp AI agent that has achieved 99.7% uptime over six months using a multi-provider AI chain with zero inference cost. The system, which serves businesses across 20 industries, employs a four-provider failover chain (Groq, Cerebras, SambaNova, Mistral) and a tool dispatcher with an autonomy gate to handle actions like reservations and inventory checks. The developer emphasizes that production AI is 10% model quality and 90% engineering, sharing lessons on WhatsApp-specific constraints such as session contention and message ordering.

Everyone demos a WhatsApp chatbot. Few run one in production with real customers sending real messages 24/7. After 18 months of running SARA — an open-source WhatsApp AI agent serving businesses across 20 industries — here's what we learned about architecture that survives contact with reality. The numbers are simple: But WhatsApp is NOT just another chat channel. It has unique constraints that break naive implementations. WhatsApp WAHA → Bridge :3008 → SARA API :3006 → AI Provider Chain → Tool Dispatcher ↓ Groq → Cerebras → SambaNova → Mistral Single-provider AI is a production risk. We use a 4-provider chain: Primary: Groq fastest, free tier ↓ fail Fallback 1: Cerebras ↓ fail Fallback 2: SambaNova ↓ fail Fallback 3: Mistral paid, always works Each provider gets 2 retries with exponential backoff before failover. Result: 99.7% uptime over 6 months with $0 inference cost free tiers . SARA doesn't just answer questions. She executes actions: create reservation — books a table with date normalization "domani alle 8" → 2026-08-10T20:00 check inventory — queries stock levels generate invoice — creates a PDF from database records schedule appointment — manages calendar slotsThe dispatcher maps 30+ tools to handlers with an autonomy gate: User message → Intent classification → Risk assessment → Tool execution ↓ Low risk: execute immediately Medium: execute + notify owner High: ask for confirmation first You do NOT want your AI agent booking a catering order for 500 people without human approval. Messages contain names, phone numbers, addresses. Our pipeline: WhatsApp doesn't have "sessions" — it's just a stream of messages. We manage context with: SARA runs on a single VPS 4 vCPU, 8GB RAM : | Component | Resource | |---|---| | WAHA WhatsApp Web | ~500MB RAM | | Bridge service | ~50MB | | SARA API | ~200MB | | PostgreSQL + pgvector | ~2GB | | Total | ~3GB | No GPU needed — inference is offloaded to cloud providers Groq, etc. . WhatsApp session contention — running two instances with the same number = instant logout for both. We learned this the hard way. Date parsing across languages — "dopodomani" Italian for "day after tomorrow" + timezone handling + business hours awareness. This alone took weeks. Message ordering — WhatsApp doesn't guarantee delivery order. Our bridge queues and re-orders by timestamp. SARA is AGPL-3.0 on GitHub: github.com/Alessandro114/sara https://github.com/Alessandro114/sara Self-host it, extend it, build your own vertical agent on top. Cloud-only features multi-tenant, white-label, analytics stay in the commercial version. The 20 industry-specific agent definitions are also open source: scala-agent-definitions https://github.com/Alessandro114/scala-agent-definitions Apache-2.0 . Running AI in production is 10% model quality and 90% engineering. Follow for more war stories.