Shako

Provides a canvas-based interface for constructing business process automation workflows without code, using a node-and-edge graph model where users connect predefined action blocks (triggers, conditions, data transforms, API calls) to define sequential or branching execution paths. The builder likely uses a state machine or DAG (directed acyclic graph) pattern to validate workflow topology and prevent circular dependencies, with real-time preview of execution flow.

Enables creation of customer-facing conversational agents through a visual dialogue tree or intent-matching system, where users define conversation paths, user intents, and bot responses without coding. The system likely uses NLP intent classification (possibly via transformer models or rule-based matching) to route user messages to appropriate response branches, with support for context persistence across conversation turns and integration with backend workflows.

Provides mechanisms for handling workflow failures, including retry policies (exponential backoff, fixed delays), error routing (alternative paths on failure), and error notifications. When a workflow step fails, the system can automatically retry the step with configurable delays and maximum attempts, or route execution to an error handling path for manual intervention or alternative processing. Error details are logged for debugging.

Enables scheduling of workflows to run at specific times or intervals using cron expressions or a visual schedule builder (daily, weekly, monthly, custom intervals). The system maintains a scheduler that evaluates trigger conditions at specified times and initiates workflow execution. Scheduled workflows may support timezone configuration and can be paused, resumed, or modified without redeployment.

Implements a publish-subscribe or event-driven architecture where workflows are initiated by predefined triggers (scheduled times, incoming webhooks, form submissions, API calls, or manual invocation). The system routes incoming events to matching workflows based on trigger conditions, executes the workflow DAG sequentially or in parallel where applicable, and manages execution state and error handling. Likely uses a job queue or message broker pattern to decouple trigger reception from workflow execution.

Provides built-in data transformation capabilities within workflow steps, allowing users to map, filter, aggregate, or restructure data flowing between workflow nodes without external ETL tools. Likely supports JSON path expressions, template literals, or a visual field-mapping interface to extract and reshape data from API responses, form submissions, or previous workflow steps. May include basic functions for string manipulation, date formatting, and conditional value assignment.

Enables workflows to call external APIs, webhooks, or SaaS services through HTTP-based action blocks that support GET, POST, PUT, DELETE methods with configurable headers, authentication (API keys, OAuth, basic auth), request bodies, and response parsing. The system likely maintains a library of pre-configured integrations for common services (email, SMS, CRM, payment processors) with simplified configuration, while also supporting generic HTTP calls for custom integrations. Response handling includes status code checking, JSON parsing, and error routing.

Provides visibility into workflow execution history, including execution timestamps, status (success/failure), duration, input/output data, and error messages. The system likely stores execution logs in a time-series database or log aggregation system, with a dashboard or UI for querying and filtering execution history. May include basic alerting for failed executions or performance anomalies, though advanced monitoring features are likely limited on the free tier.

Allows workflows to branch execution paths based on conditional logic (if-then-else, switch statements) evaluated against workflow data, previous step outputs, or external data sources. Conditions are likely defined through a visual rule builder or expression language supporting comparisons (equals, greater than, contains), logical operators (AND, OR, NOT), and nested conditions. The system routes execution to different workflow paths based on condition evaluation, enabling complex business logic without code.

Enables creation of web forms that collect user input and automatically trigger workflows with the collected data as input. Forms likely support various field types (text, email, phone, dropdown, checkbox, file upload) with validation rules, and submission data is passed directly to workflow steps for processing. The system may provide embeddable form widgets or hosted form pages, with optional styling and branding customization.

Provides pre-built workflow templates for common business processes (customer onboarding, lead qualification, invoice processing, support ticket routing) that users can clone and customize rather than building from scratch. Templates likely include pre-configured triggers, actions, and conditional logic specific to the use case, with placeholder values for customization. The system may support creating and sharing custom templates within teams or across the platform.

Supports execution of workflows containing multiple steps that can run sequentially (one after another) or in parallel (simultaneously) depending on workflow design. The system manages dependencies between steps, ensuring data flows correctly from one step to the next, and handles parallel execution coordination and result aggregation. Likely uses a DAG execution engine with topological sorting to determine execution order and manage state across steps.