Migrate from Timefold to Freshsales

Timefold constraint model has no native equivalent in Freshsales — optimization logic does not migrate

Timefold solves constraint-satisfaction problems: it takes problem facts (resources, locations, time windows) and planning entities (visits, shifts) and returns an optimized assignment with a scored solution. Freshsales has no concept of constraint solvers, score directors, or planning variables — it is a transactional CRM. FlitStack migrates Timefold's output data (the optimized schedule) into Freshsales custom objects, but the constraint definitions, score weights, and solver configuration must be documented from Timefold and rebuilt as Freshsales workflow rules, assignment logic, or manual processes. Teams expecting 'scheduling intelligence' in Freshsales will find that the migrated data represents a static schedule snapshot, not an active optimization engine. The constraint model cannot be imported as a running system — it must be re-implemented or accepted as historical data only.

Custom objects are gated to Freshsales Enterprise — Growth and Pro plans cannot receive scheduling data

Freshsales custom objects are only available on the Enterprise plan (formerly Forest tier). Timefold's scheduling problems, schedule records, and constraint configurations require custom objects to map faithfully. On Growth ($9/user/month) or Pro ($39/user/month) plans, these records have no place to land — they can only be stored as custom fields on Contact or Account, which severely limits the relational depth of the migration. We flag the plan level before schema setup and advise Enterprise qualification if the scheduling dataset is core to operations. This is a billing-model surprise: Freshsales advertises custom fields on Pro, but custom objects — the structural equivalent of Timefold's entities — are Enterprise-only.

Freshsales API rate limits throttle migration throughput on lower plans — large datasets require batched runs

Freshsales API rate limits are tiered: Growth allows 1,000 requests/hour, Pro allows 2,000/hour, and Enterprise allows 5,000/hour. Timefold planning datasets can contain tens of thousands of records per planning run (employee × visit combinations). FlitStack implements request-level throttling and batched upserts to stay within the per-hour limit on Growth and Pro plans, which extends migration run windows for large datasets. We surface the rate-limit constraint during discovery and calibrate batch sizes to the destination plan tier. Exceeding the rate limit returns HTTP 429 responses and triggers automatic retry with backoff.

Timefold records without email addresses cannot create Freshsales Contacts — gap between planning IDs and CRM records

Freshsales requires a valid email address to create Contact records via the API. Timefold employee and resource records often lack email addresses, particularly for anonymous field resources or third-party contractors. Records without email are flagged before migration and routed to a manual review queue: either the email is added to Timefold and the migration re-run, or the record is created as a Freshsales Contact with a placeholder email and a note flagging the gap. This is a data quality issue that requires human judgment and cannot be auto-resolved. We surface it in the pre-migration audit report.

Timefold solver scores and constraint-violation data are opaque without the Java domain model

Timefold solver scores (e.g., '0hard/-257medium/-6119520soft') reference constraint weights defined in the solver's Java domain model — the score alone does not explain which specific constraints were penalized or why. When this score is migrated to Freshsales Score__c on the Scheduling_Problem__c record, it provides a reference point but not an actionable explanation. Teams that relied on Timefold's score breakdown for root-cause analysis of schedule quality will need to export the full solver output JSON and maintain it as a reference artifact outside Freshsales. We include this export as part of the migration artifact package.