1 /

The declarative, AI-native platform for the modern enterprise

The BPA Approach

A different path through the AI shift

Composes, not black-box magic

AI assembles solutions from declarative building blocks you already understand.

You stay in control

Every decision is readable, editable, versioned, auditable. No black boxes.

AI-configurable, not dependent

Run with AI acceleration, or configure manually. Your team is never locked out.

Compared To

How BPA is different

Category Great for Typical trade-off BPA's approach
AI code generators Rapid prototyping and MVPs Output is code your team then owns, hosts, and maintains Declarative configuration running on the platform — nothing extra to maintain
AI-enabled low-code Internal tools and departmental apps AI typically sits as a helper on top of the builder AI is a participant in the running workflow — same tools, same data, same guardrails
Enterprise low-code & BPM Large regulated workflows with established governance Mature and proven; newer AI capabilities often added as premium layers Declarative and AI-native from the foundation, with bring-your-own AI via MCP
Enterprise AI agent platforms Customers standardised on one vendor's ecosystem AI agents are typically tied to the hosting vendor's data and AI stack Vendor-independent — connect your own AI via MCP; keep your data in one governed layer
Orchestration

Where enterprise work actually happens

Systems, people, and AI — participating in the same flow.

Connect everything

Your systems, partners, and customers — unified through first-class integrations.

AI joins the process

AI acts inside the workflow, where it adds value — humans in the loop for judgement.

Governed end-to-end

Every system call, AI decision, and human action in one audit trail.

Capabilities

One platform, eight capabilities

Entities
Unified data with validation & security
Forms
Rich field types, conditional logic
Flows
Visual flow designer, actions, SLAs
Apps
Composable screens & dashboards
AI Agents
Declarative, guarded, draft-safe
Bundles
Export, version, promote across environments
Security
Row & field rules, unified audit
Analytics
Process metrics, AI cost & usage
Ecosystem

BPA in your world

Users & Teams Web · Apps · AI Sidebar Partners & Customers Forms · Portals · API Business Process Automate BUILD Entities Forms Flows Apps Bundles Declarations AI FRAMEWORK Declarative Agents Knowledge Packs Tool Registry Solution Builder TRUST & GOVERNANCE ISO 27001 Row & Field Rules RBAC Audit Trail MFA REST API · OpenAPI · Webhooks · Events · MCP Server CONNECT YOUR OWN AI OR USE BPA'S — Same tools · Same guardrails · Same audit trail MCP Clients Claude Desktop · Cursor · Windsurf · Custom Your Systems ERP · CRM · Finance · HRIS · APIs Events & Webhooks Outbound notifications · Subscriptions SaaS (Azure Managed) Dedicated (Azure Tenant) Customer VPC (Azure) POWERED BY MICROSOFT AZURE
BPA ecosystem — platform, people, and integrations
Security

Security built into every layer

Authentication Paths Google OIDC OpenID Connect Microsoft Entra Azure AD / OIDC Local Auth Argon2 Hashing JWT Engine Access + Refresh MFA / TOTP Authorisation Layers (evaluated per request) System Roles Platform-wide access for authorised administrators Team Roles Tenant-scoped owners, admins, and members App Roles Per-application permissions, screen access Entity Security Policies Row-level access rules, field-level masking ABAC Conditions + Time-Based Access Attribute matching, temporary grants Token Unified Audit Log — auth · entity · security · AI
Figure 2 — Authentication flow and layered authorisation, ISO 27001 certified
AI

AI-native, not AI-bolted-on

AI Framework — Layered Architecture PROVIDER LAYER OpenAI Anthropic Claude Azure OpenAI Google Any LLM provider KNOWLEDGE LAYER Forms Pack .md knowledge Process Pack .md knowledge Entity Pack .md knowledge MEMORY LAYER Session Conversation Agent Persistent Team Org-wide shared INTEGRATION LAYER — Typed Tool Registry BPA Tools (Entity/Form/Flow) Email Tools HTTP Tools Flow Tools Custom Tools Declarative Agent Runtime System Prompt + Tools + Knowledge + Guardrails — loaded from DB at runtime Configurable iteration, token, entity, and action limits per agent Five AI Surfaces — My AI · Sidebar · Solution Builder · AI Studio · MCP Same tools, knowledge, and guardrails across every surface
Figure 3 — Layered AI framework, declarative agents, draft-safe writes
Your AI, Your Choice

Use BPA's AI — or connect your own

Built-in

BPA's own AI framework

  • Surfaces: My AI · AI Sidebar · Solution Builder · AI Studio
  • Managed providers: OpenAI · Claude · Azure · Google and more
  • Supervised agent squads, draft-safe writes, full guardrails
No setup. Works out of the box.
Your choice

Connect Your Own AI

  • Claude Desktop · Cursor · Windsurf · Custom clients
  • Connected via the standard MCP protocol
  • Full access to BPA's tools, data, and knowledge packs
Use your existing AI investments and skills.
Solution Builder

From brief to running app — supervised by humans

Four squads of specialised agents. Every document, design, and decision is yours to review before it moves forward.

01
Discovery
Discovery Agent
BRD Writer
Produces
BRD
02
Design
Solution Design Agent
Produces
Solution Design
03
Build
Planner · Entity · Form
Flow · App · Verifier
Produces
Working App
04
Test
Test Planner
Test Executor
Produces
Test Plan + Report
Portability

Move, version, and promote with confidence

Integration Architecture EXTERNAL SYSTEMS Third-Party APIs Webhook Subscribers CI/CD Pipelines Custom Applications Business Process Automate REST API OpenAPI Docs Webhooks (Out) Event Catalog MCP Server Decl. Bundles PORTABILITY Export Bundles (JSON) Import with Merge Version History Environment Promotion All interactions authenticated — every event unified in one audit trail
Declaration bundles, environment promotion, and no vendor lock-in
Data Model

One unified record store

Entity-Centric Data Architecture Form Submissions Rich field types Workflow Actions Typed action library App Operations CRUD via rules AI Agent Actions Tool execution REST API (Direct) Open REST API Entity Definitions (Schema + Validation + Security Policies + Relationships) Field types | Computed fields | Conditional requirements | Transforms | Unique constraints entity_records (Unified Data Store) Single collection — partitioned by entity_key + team_id — versioned + audited COMMON DATA MODEL (CDM) Core CRM Helpdesk Sales Reserved entity keys across domain packs
Figure 4 — One unified entity_records store with CDM domain packs
A Day in the Life

A new hire joins — productive on day one

Candidate, manager, HR, compliance, and IT — all working in one flow. No hand-offs that leak.

01
Candidate
Submit
New hire completes onboarding forms through your portal
02
AI
Assist
Extracts IDs, tax forms, and right-to-work documents
03
Compliance
Verify
Background and right-to-work checks run inline via your providers
04
Manager & HR
Approve
Review the AI-prepared summary and approve the new hire
05
IT & Systems
Provision
Identity, email, laptop, payroll, and access groups — one audit trail
Week 2 → day 1. Every step configured, supervised, and auditable — with or without AI.
Where it works

Any regulated process. Any team. Any vertical.

BPA's declarative pattern works wherever enterprise work needs structure, audit, and AI assistance.

Banking & Financial Services Real Estate & Funds Healthcare Government Education Insurance Legal Manufacturing Utilities Retail Logistics Professional Services
The pattern is the same. The configuration changes.
Microsoft Azure hosted · ISO 27001 certified · Multi-tenant · Region residency (AU · EU · US · UK · SG)
Next Steps

Let's explore what BPA can do for you

  • 1. Discovery call — we learn about your business and current challenges.
  • 2. Walkthrough — we show you how BPA applies to your specific scenario.
  • 3. Evaluation — hands-on access to explore BPA with your team.
  • 4. Next steps — we agree on scope, timeline, and how to move forward.