Owned productSaaSWhatsApp Business APIShared inboxCampaignsChatbot automationAI knowledge replies
Xenwap: Turning WhatsApp conversations into a managed growth workflow
Xenwap is Xenova’s AI-powered WhatsApp growth platform. It gives teams a shared place to manage customer conversations, run approved campaigns, automate routine journeys, and answer questions using business knowledge.
The problem
WhatsApp is often where customers respond fastest, but unmanaged business messaging creates risk and friction:
- Conversations sit on individual phones with no shared ownership.
- Teams cannot see who is responding or what has already been promised.
- Repetitive questions consume staff time.
- Campaigns are separated from replies and follow-up.
- Customer context lives in other systems.
- Generic AI answers can be wrong, off-brand, or impossible to audit.
- Consent, templates, delivery states, and platform rules are easy to mishandle.
The product challenge was to turn a high-volume communications channel into an accountable team workflow without losing the immediacy that makes WhatsApp valuable.
The solution
Xenwap combines four connected capabilities.
Shared team inbox
Teams can work from a common queue, assign ownership, view conversation context, and reduce duplicate or missed replies.
Campaign management
Approved outbound messages can be organised as campaigns with audience logic, templates, and delivery status. Replies return to the operational inbox instead of becoming a separate dead end.
Chatbot automation
Rules and conversational steps handle repeatable requests, collect information, and route exceptions to a person.
AI knowledge replies
The assistant uses approved organisational knowledge to propose or deliver relevant answers. The safe product pattern includes source governance, confidence boundaries, escalation, and conversation logs rather than unconstrained generation.
A representative workflow
- 01A business connects its verified WhatsApp setup and configures authorised team access.
- 02Knowledge sources, message templates, routing rules, and escalation paths are defined.
- 03An inbound message reaches the shared inbox.
- 04Automation classifies the request and either follows an approved flow, proposes a grounded answer, or assigns it to a person.
- 05The team member sees the conversation and takes ownership when judgment is needed.
- 06Delivery, read, reply, assignment, and resolution events remain visible for follow-up and reporting.
Architecture decisions
Model messaging as asynchronous events
Webhook delivery, retries, out-of-order events, duplicate notifications, media processing, and provider errors are normal conditions. Reliable messaging software must be idempotent, observable, and designed for eventual state changes.
Separate channel policy from customer workflow
WhatsApp template categories, conversation windows, consent, and account-quality requirements evolve. Keeping platform-specific rules isolated from broader campaign and support logic reduces the cost of change.
Design multi-user ownership explicitly
A shared inbox is not simply a list of messages. It needs assignment, concurrency handling, permissions, unread state, internal context, and an auditable record of who acted.
Ground AI in approved knowledge
Knowledge ingestion needs access rules, chunking and indexing choices, update and deletion paths, source visibility where appropriate, evaluation questions, and a fallback when confidence is insufficient. Human takeover must be easy.
Protect customers from unsafe campaigns
Audience consent, opt-out handling, rate controls, template status, frequency, and error monitoring belong in product design. Growth cannot depend on behaviour that threatens the customer's channel access.
Technology and delivery
vuenuxt.jsnode.jsmysql
Xenwap’s interface is built in Vue with Nuxt.js, backed by Node.js services and MySQL. Node.js suits the workload: a messaging platform spends most of its time waiting on network events (webhooks arriving, messages sending, statuses updating), and an event-driven runtime handles that concurrency naturally.
Conceptually, the platform moves a message from Meta webhooks through ingestion and validation, into an event queue, through the conversation and automation layer, and out to the inbox and campaign interfaces, with analytics and monitoring alongside.
meta.webhook → ingest.validate → event.queue → conversation.service → automation / ai → inbox · campaigns
Conceptual flow: exact production design published after engineering review.
Infrastructure specifics (hosting, queueing, and the exact integration arrangement) are shared in technical discussions rather than published here.
Security and operations
Multi-tenant messaging demands care: tenant isolation, role-based access, credential storage, webhook validation, audit events, retention and deletion, backups, and AI-provider data handling. We document and verify each of these per deployment rather than publishing generic security language.
Results
Xenwap is owned product proof that Xenova understands the complete WhatsApp workflow: integration, team operations, campaigns, automation, AI knowledge, and ongoing platform constraints.
We make no numerical outcome claims here until measured figures are approved with their period and source. A working product tour, verified architecture detail, and a named product owner are the initial evidence.
What this proves for clients
Xenwap demonstrates experience with integration-heavy SaaS, asynchronous systems, multi-user permissions, real-time work queues, business automation, and constrained AI communication. Those patterns transfer directly to support platforms, commerce operations, lead management, internal service desks, and notification-heavy products.
See our software development services →Previous case study: Skoo →