Decisions

What it is

Decisions is an intelligent process orchestration platform for enterprise teams. Its site says it brings workflows, rules, AI orchestration and governance together in one platform, "instead of stitching together separate tools", so an organization can build, automate and control how work runs in one place. It is a low-code builder aimed at business and technical teams working side by side, and the company says changes should not depend on developer backlogs. The home page frames the product as the layer that fills gaps around core systems: core systems were not built for every workflow, decision, exception, approval or AI action, and Decisions adds the automation, orchestration and governance on top. The About page says Chief Technology Officer Heath Oderman co-founded Decisions and has led its development and delivery teams since 2010, and that Chief Customer Officer Brian Reale previously cofounded ProcessMaker and ran it before its merger with Decisions.

What you build and how

The platform overview lists six core capabilities: AI orchestration, an enterprise rules engine, workflow and process automation, application development, process intelligence and integrations. The pages treat them as one system rather than separate products.

Workflows and process automation. You design processes in a drag-and-drop visual studio, connecting steps, branching on logic and managing exceptions. The company says the range runs from single-task workflows (approvals, notifications and form-driven actions) to multi-system processes with cross-team coordination, conditional logic and system integrations. Its workflow page lists what you can build: approvals, requests, case workflows, onboarding workflows, exception handling, system orchestration and industry-specific workflows. Visual workflow design and reusable rules mean teams can change logic, routing, approvals and steps "without digging through custom code". The company contrasts this with basic workflow tools that route tasks: Decisions combines workflow automation with rules, integrations, orchestration, governance and agent coordination.

The rules engine. The company calls rules the foundation rather than a feature: every workflow, form, integration and AI interaction is powered by centralized, visual logic you can see, test, trace and change. A rules engine evaluates business logic and returns an outcome such as approve, reject, route, calculate, escalate or request more information. The site draws the line between the two engines this way: a workflow engine moves work between steps, a rules engine decides what should happen at each step, and Decisions runs both together, so rules drive workflow branches and workflows invoke rules. There is no handoff between a rules engine and a separate BPM tool. The company says the engine processes billions of evaluations per hour. The agentic page names statement rules, truth tables and matrix rules as rule types.

Agents and the agent builder. This is where the product touches chatbot territory. The home page promises to "build and manage first- and third-party agents at scale with visual tools to design agent actions", backed by a rules engine for governed intelligent decisioning and process visibility into granular agent workflows. The agentic orchestration page describes the problem it solves, AI sprawl without guardrails, where every team builds agents, no one coordinates them and logic is scattered and impossible to version or roll back. Its answer is a control plane: you define exactly what each agent is authorized to do and what it is not, using statement rules, truth tables and matrix rules, so that agent behavior stays inside deterministic boundaries. The workflow page adds that agents can take part in workflows under rules, approvals, audit trails and controls. The pages describe agents as governed actors inside business processes: in the trial you build first-party agents or add third-party ones and orchestrate them alongside people, rules and systems. On the customer page, the National University of Colombia story is summed up as "100% fully automated customer support". The company also has a Buyer's Guide on process and AI orchestration platforms.

Application development. The platform includes tooling to design, test and deploy applications, forms, dashboards and more. The company says that if you can define it, you can build it, including an application no vendor ships out of the box.

Process intelligence. This module maps tasks, processes and AI agent activities, so you can see what is moving, what is stalled and where to optimize.

Integrations. The company lists hundreds of out-of-the-box connectors and a suite of integrations. Its financial services page names core banking, loan origination (LOS), loan management (LMS), mortgage platforms, CRM, third-party systems and proprietary tools.

What the pages say

Outcome figures. The home page lists 70% faster rules changes, 70% less manual work time and 40% lower cost of ownership under the heading "faster, more cost-effective orchestration". The customer page attributes its outcomes to named customers: a 75% reduction in human errors at Maury, Donnelly & Parr Inc., a $300,000 increase to the bottom line at Miami Parking Authority, 3X data volume handled at KnowLedger, and healthy days and quality-of-life metrics improved by 45%-60% at Vantage Health in oncology care. Kaiser Permanente and Fire Recovery appear as case studies too.

Plans and trial. The pricing page compares three tiers named Foundation, Growth and Enterprise across features such as the low-code design studio, agentic AI and agent builder, and the agent and business rules engine, with segment views for Enterprise, Mid-Market and Higher Ed. The site publishes no prices. The company says it is not built around restrictive per-user pricing, that you can start small and expand later, and that a custom plan with custom limits, advanced security and dedicated support is available on request. The free trial runs 14 days: a guided SaaS trial, fully managed in the cloud, where the company spins up the instance, pre-loads sample apps and training, and gives access to its human support team. The trial walks you through installing a template from the sample app library, designing a workflow on the canvas, adding decision tables, and testing and deploying.

Security and deployment. The pricing page states "SOC 2, HITRUST, HIPAA, ISO 27001, PCI DSS certified. Built for regulated industries." It also says Decisions runs in the cloud, hybrid or on-premise, and the process automation page adds "Any AI model". The company says every rule and decision is auditable by design, that compliance teams can update policy in minutes with full version history regulators can read, and it describes AI as something to govern, "not just deploy".

Support and reviews. The customer page lists professional services, support through documentation, forums, in-platform AI guidance, direct technical support and a Customer Success Manager, and training through live events, self-paced courses and certifications. The trial page adds in-app chat. The company says it is rated 4.6 out of 5 by verified customers on Gartner Peer Insights, G2 and Capterra. One quoted customer, Renee Pederson of Mid First Bank, says: "You can actually see the steps and follow the process."

Industry. The financial services page positions Decisions above core banking and LOS systems, which store data and process transactions, to drive credit decisions, compliance reviews and approval chains those systems do not handle. Insurance, manufacturing, healthcare, higher education and the public sector have their own solution pages.

Who it suits and who it does not

Decisions suits a large organization, particularly in financial services, that wants rules, workflows and AI agents under one governed roof. A team with core systems it cannot easily change, and with approvals, exceptions and credit or compliance decisions living around them, is the clearest fit. It also suits a team starting its first agent projects that wants audit trails and guardrails before agents multiply. The no per-user model helps when many people touch the same process.

It is more than a founder or a small team needs. The pricing page calls it a premium platform and plans are tiered and customized without a published price, so a team that wants to test an idea cheaply has the 14-day trial and then a sales conversation. It is not a website builder, a store builder or a lightweight chatbot tool. A regulated team that needs on-premise deployment or named certifications will find both stated on the pricing page.