Migrate from Certainly to Freshdesk

Certainly's dialogue trees and decision branches map to Freshdesk's ticket creation rules, time-triggered automations, and routing policies. We document every Flow's conditional logic, branch conditions, and fallback targets as a written handoff inventory. Freshdesk's hourly trigger limitation (processes tickets updated within the last 30 days only) means certain time-sensitive routing from Certainly may need adjustment in the destination.

Intent names and their associated training utterances export from Certainly's NLU model. We preserve the utterance set as a structured archive for retraining in Freshdesk's Freddy AI. The trained model accuracy does not transfer because intent classification depends on the destination platform's NLU engine. Retraining with the exported utterances is required before bot classification quality matches pre-migration levels.

Custom entities and slot types from Certainly — entity names, value lists, regex patterns, and slot definitions — require field-level mapping. Simple value-list entities map to Freshdesk dropdown custom fields. Regex or pattern-based entities are documented for manual recreation in Freshdesk since Freshdesk's custom object fields support text, number, decimal, date, and multi-select but not native regex validation without additional configuration.

Static response texts, rich message templates, and conditional response branches export as structured data and map directly to Freshdesk's Canned Responses and Ticket Templates. Rich formatting, buttons, and quick-reply structures are preserved where the destination supports equivalent message types. Response branching logic is captured in the Flow inventory document for manual rebuild in Freshdesk's automation builder.

Certainly's Zendesk connector configuration — ticket creation triggers, field mappings, and handoff rules — does not export as migration-ready code. We document the existing connector settings during discovery, capturing ticket creation logic, field mapping pairs, and routing conditions. The customer's admin recreates these in Freshdesk using Freshdesk's native integrations or the Freshdesk API. This is a manual rebuild step, not a data migration.

Historical chat transcripts export from Certainly as structured data with session IDs, timestamps, agent attribution, and resolution status. Freshdesk does not have a native bot-conversation log object; session data maps to Freshdesk ticket threads where applicable. For sessions that do not map to a ticket context, we archive logs as JSON or CSV alongside the primary migration to preserve historical performance data.

Bot performance metrics, intent accuracy scores, conversation volume trends, and resolution summaries export as flat files. Freshdesk's reporting engine operates on ticket-level metrics and does not ingest bot-conversation analytics natively. We archive all analytics exports for import into the customer's BI tool or for manual reference.

Agent IDs, team assignments, and role-based access structures in Certainly's routing rules map to Freshdesk Agents and Groups. We match by email address and flag any role mappings that require manual recreation in Freshdesk's permission structure. Freshdesk's agent roles (Agents, Ticket Agents, Admins) differ from Certainly's enterprise RBAC model and may require scope adjustment during configuration.