Migrate from Mava to HubSpot Service Hub

Mava Conversations map 1:1 to HubSpot Tickets. The conversation ID is preserved as a custom property mava_conversation_id__c on the ticket for audit traceability. Channel source (Discord, Telegram, or web) migrates as a single-select custom property ticket_channel__c. Message history and participant metadata transfer as threaded replies within the ticket conversation pane via HubSpot's Tickets API. Ticket status (open, pending, resolved, closed) maps from Mava's conversation state field.

Mava community members linked to Discord user IDs or Telegram user IDs without verified email addresses present a resolution challenge. We extract the full member roster and attempt email-based matching against any web session data captured in Mava. Members with no resolvable email are migrated as HubSpot Contacts with a mava_platform_id__c custom property storing the original Discord or Telegram ID and a mava_identity_unresolved__c flag set to true, so the customer's admin can enrich records post-migration. Members with verified emails map 1:1 to HubSpot Contacts.

Mava Agents map directly to HubSpot Users with Service Hub seat access. We resolve by email match. The agent role (admin, agent) maps to the HubSpot user permission level. The customer's admin provisions HubSpot Service Hub seats before migration begins so that OwnerId references are satisfied at import time.

Mava Teams map to HubSpot Teams. Team names and member lists migrate as-is. Team-based routing in Mava maps to HubSpot ticket assignment rules tied to the team object. If Mava teams are channel-specific (e.g., a Discord team and a Telegram team), we consolidate them into HubSpot Teams and note the channel association in the team description field for the admin to assign routing rules in HubSpot.

Mava conversation tags migrate as HubSpot ticket properties. Tags are preserved as key-value strings in the mava_tags__c multi-select custom field on the ticket. If tag volume exceeds HubSpot's multi-select picklist limits, tags migrate as individual Topic assignments or as a comma-separated text field, with the customer choosing the strategy during scoping.

Mava AI bot intent catalogs are configured per channel with intent rules, automated responses, and routing conditions. HubSpot Service Hub has no equivalent bot-intent object — Breeze AI Agent uses knowledge-base retrieval rather than intent-based routing. We extract the full bot intent catalog across all channels (Discord, Telegram, web) as a JSON manifest and deliver it as part of the migration documentation package. Rebuilding intent rules in HubSpot requires manual configuration of Breeze AI Agent knowledge sources and is outside the migration scope.

Mava webhook integrations are customer-defined with no documented standard payload schema, meaning payload structures vary significantly per customer. We extract all webhook endpoint URLs, trigger conditions, and associated conversation IDs as a manifest. HubSpot Service Hub supports native webhooks via Workflows (outbound webhook action) and the HubSpot API for inbound webhook delivery. Payload transformation requires manual mapping and is documented for the customer's developer to implement post-migration.

Mava SLA configurations (first-response and resolution time targets per channel or team) map to HubSpot Service Hub SLA Policies, which support targets on first response time and next time to reply. We document each Mava SLA rule (target time, business hours, priority) and the associated team or channel, then configure equivalent SLA Policies in HubSpot during migration. SLA policy activation requires the Professional or Enterprise tier of HubSpot Service Hub.

File attachments within Mava conversations migrate as HubSpot file attachments linked to the corresponding ticket. We use HubSpot's Files API to upload each attachment and link it via the ticket hs_attachment_ids property. Large files (over 16 MB) exceeding HubSpot's file size limit are flagged in the migration manifest for manual handling.

Mava conversations carry channel source metadata indicating whether the thread originated on Discord, Telegram, or web. HubSpot Service Hub does not have a native channel source field on tickets. We create a single-select custom property ticket_channel__c on the Ticket object and populate it with the source value during migration. This preserves channel attribution for reporting without requiring agents to infer the source from message timestamps or content.

We store the original Mava conversation identifier as a string custom property mava_conversation_id__c on each migrated HubSpot Ticket. This creates a permanent audit link between the destination record and the source record, enabling reconciliation and supporting any future cross-system reporting the customer requires.

Individual messages within a Mava Conversation thread map to reply entries in the HubSpot Ticket conversation pane. We preserve author identity (agent or contact), timestamp, and message body. Discord message threading (reply chains) and Telegram message groups transfer as sequential reply entries ordered by the original timestamp. Rich formatting, embedded media URLs, and emoji reactions from Discord and Telegram do not preserve formatting; plain message content and links transfer as-is.