How DRAK turns a business-analyst's process model into deployed, multi-chain smart contracts — no Solidity required.
A blockchain application is expressed by a business analyst using BPMN (control flow) and DMN (decision rules). DRAK transforms the model into a multi-modal Discrete-Event Hierarchical State Machine (DE-HSM) — time as discrete events, functionality as hierarchical state machines — and then into the methods of one or more smart contracts. Synchronization, cryptographic security and the transaction mechanism are generated automatically, so developers code only isolated task bodies.
Using SESE subgraphs of the DE-HSM model, DRAK finds patterns that are isolated from other concurrent activity and can run on a cheaper sidechain. Each pattern's mainchain cost is compared with its sidechain cost plus cross-chain overhead; cheaper patterns are deployed off-chain.
Long-term, collaborative activities spanning many methods are supported as nested trade (sub)transactions with Atomicity, Consistency, Isolation and Durability enforced by a generated mechanism — sub-transactions can be packaged as separate contracts on different chains for privacy.
When an unanticipated event prevents completion (e.g. a flood washes out a rail line), the failure is localised to the innermost sub-transaction; the modeller amends that BPMN fragment under pre/post-repair conditions, a new contract version is generated and deployed, the API is repointed, and execution resumes — preserving previously completed activities.
Payments are not modelled in BPMN; generic payment tasks are bound at runtime to pre-deployed payment services from a repository (on-chain native, cross-chain stablecoin USDC/USDT, off-chain card / wallet), with settlement-netting that aggregates many payments into a single transfer.