Migrate from Mava to Intercom

Mava Conversations map 1:1 to Intercom Tickets. Each Conversation's message history, timestamps, channel source (Discord, Telegram, or web), and participant list migrate into Intercom's Ticket object with the messenger thread preserved underneath. The Mava conversation ID is stored as external_id on the Intercom Ticket for traceability. Ticket title is derived from the first message or a subject line if one exists; otherwise we flag for customer review before migration.

Mava community members associated with a verified email address map directly to Intercom Contact. Members linked only to Discord user ID or Telegram user ID require identity resolution before import: we attempt to match against known email addresses in the Mava data export, and records without email go to a reconciliation queue. The customer reviews unresolved identities and either enriches them with email or approves mapping to anonymous contacts. The original platform identity (Discord/Telegram) is stored as a custom attribute on the Contact.

Mava Agents (support team members with name, email, role, and assignment rules) map 1:1 to Intercom Admin. We resolve each Agent by email match against the destination Intercom workspace. If a matching Admin does not exist, we create a placeholder or queue it for provisioning before migration. Agent role and assignment preferences are documented separately as Intercom Inbox configuration steps since Mava's role model does not map directly to Intercom's Admin and Agent permission sets.

Mava Teams (names and member lists used for routing) map 1:1 to Intercom Teams. The team name migrates as-is, and Agent membership is mapped by resolving each Mava Agent to their Intercom Admin counterpart. Inbox routing in Intercom is configured post-migration using the team assignments we document in the migration manifest.

Mava conversation Tags (simple key-value strings used for categorization) migrate directly to Intercom Tags. Tag assignments on individual Conversations migrate as Intercom Tag associations on the corresponding Ticket. Tag counts are preserved exactly; we do not merge or deduplicate tags unless explicitly requested by the customer during scoping.

Mava SLA configurations (first-response and resolution time targets tied to channels or teams) map to Intercom SLA Policies. We extract SLA rules including target times, business hours definitions, and channel scope, then configure these as Intercom SLA Policies tied to the appropriate Inboxes. Any pause conditions or escalation rules in Mava are documented for manual recreation in Intercom since SLA pause logic differs between platforms.

Mava AI bot intents and automated responses configured per channel type (Discord, Telegram, or web) do not map 1:1 to Intercom's Fin AI Agent, which uses a unified Knowledge Hub and Procedures model. We extract the full Mava intent catalog across all channels, consolidate unique intents and automated responses, and deliver a written Fin AI Agent setup guide that maps each Mava bot intent to a Fin Resolution Goal or a Fin Procedure. The customer's admin rebuilds the AI layer in Intercom using this guide.

Mava webhook configurations (endpoint URLs, trigger conditions, and payload structures) are customer-defined and lack a standardized schema. We extract all webhook endpoint URLs and trigger event types and document them in the migration manifest. Payloads do not migrate; the customer maps the event triggers to Intercom's outbound webhooks or Workflow actions manually. We flag which Mava webhook triggers have no Intercom equivalent and suggest alternatives.

Mava preserves the channel source (Discord, Telegram, or web) as a participant-level metadata field. Intercom Tickets do not have a native channel-source attribute across the messenger. We create a custom Ticket attribute (e.g., original_channel__c) and populate it with the source value so that agents can see at a glance whether the conversation originated in Discord, Telegram, or the web widget without relying on the message content.

Mava SLA targets (first response time, resolution time) and current status for each conversation are stored as rule metadata rather than per-conversation fields. We extract these values and write them as custom Ticket attributes in Intercom so that SLA performance data is attached to each Ticket and visible in reporting. Business hours definitions migrate as Intercom Working Hours configurations.

Mava's AI bot automated responses contain structured Q&A pairs that function as an informal knowledge base. We extract these intent-response pairs and consolidate them into Intercom Articles organized by topic. Articles are created as Internal Articles first for agent review, then promoted to Public Articles once the customer's admin approves the content. This gives the team a Knowledge Hub foundation for Fin AI to answer questions from day one.