Interface with Twilio to interact with large audiences (10k-100k users) through WhatsApp

We are looking to develop an interface with Twilio Conversations API ([login to view URL]) to allow us to send out multiple new messages (templates) every week via WhatsApp to large audiences, and allow users to interact with our WhatsApp Business Account through message specific flows. The purpose of this is for campaigning, marketing, and fundraising initiatives of a global non-profit. We are looking to build an initial MVP to be able to handle between 1k and 100k users. The interface should be efficient and able to deliver the messages quickly and reliably, in multiple languages, and segment the list to send different messages to different audiences, sometimes in parallel in the form of A/B tests. It should subsequently be able to trigger different flows depending on user actions. It should also be able to show key statistics of user actions (clicks, user messages, unsubscribed, etc) both related to a specific campaign and overall of the entire list of users over time. The initial templates and the client-triggered flows should be customizable (we can provide a few typical flows) and multi-lingual by default (we operate in 17 different languages, although many comms happen in a subset of those lands, and sometimes just in English). The interface should be able to handle a dynamic database of users from two sources 1) existing users (or members as we call them, as we're an NGO) for which we already have their WhatsApp number and their consent. 2) users from our email list, visitors to our website, and new users coming from referrals of other users that we convert "convert" to our WhatsApp Business Account. Those should be added to the database dynamically after they sign up and consent to our comms (through interaction with our WhatsApp Business account). Also, the system should be able to handle unsubscribes. While an initial set of users will be provided as a CSV, and the initial MVP should include a database to store the dynamic list of users, the interface should already be designed to allow future integration with our own larger multi-platform database (through APIs) where all user data will be stored.