Platform
Reads. Decides. Acts.
The system of decision.
Inversiq connects documents, data, models, decision logic, workflows and human control in one governed operating layer.
Connected inputs
Inversiq operating layer
Decision record
Reading inputs into one governed decision record.
Architecture
One platform. One data model. End to end.
Evidence, models, rules, review and execution run on one governed operating layer.
The underlying engine is domain-neutral and designed to support multiple verticals. Commercial Real Estate is the first full vertical, and where the platform is applied most deeply today.
Documents & data
Documents, models, systems, records
Relationships
Entities resolved, sources reconciled
Risk & assumptions
Conflicts and gaps surfaced
Decision logic
Rules applied, rationale attached
Human review
Approval gates, accountability
Action
Workflows executed, outcomes logged
Documents & data
Documents, models, systems, records
Relationships
Entities resolved, sources reconciled
Risk & assumptions
Conflicts and gaps surfaced
Decision logic
Rules applied, rationale attached
Human review
Approval gates, accountability
Action
Workflows executed, outcomes logged
One governed operating layer — evidence to executed action.
Platform Architecture
One shared engine.
Deep vertical products.
Customer-specific configuration.
Inversiq separates reusable decision infrastructure from vertical logic and customer configuration, allowing each implementation to reflect how an organisation works without fragmenting the core platform.
Shared Engine
Execution, evidence, review, integrations, identity, audit and governance.
Vertical Layer
Domain objects, models, document types, rules, workflows and outputs.
Customer Configuration
Mappings, model structures, approval policies, terminology, templates and integrations.
Security, reliability, observability and compliance readiness across every layer.
Platform Capabilities
The capabilities behind
every governed decision.
Each capability works from the same connected sources, models, rules, permissions and audit trail.
Evidence Intelligence
Turn documents and data into structured, source-linked and reviewable evidence.
- Documents, operational data and model inputs are read into one structured record.
- Every value keeps the reference it came from, so a reviewer can go back to the source.
- Anything that cannot be confirmed is raised rather than assumed.
Raw sources
Structured evidence · source-linked
Three records source-linked · one requires confirmation before it is used.
Risk & Assumption Intelligence
Validate assumptions and surface material conflicts before they affect the decision.
- Assumptions are tested against the sources they depend on.
- Conflicts are stated in one line, with the requirement that contradicts them.
- Materiality decides what reaches a person, so review effort follows the risk.
Active assumption
Capacity is committed through 2029.
Supporting source
Commitment recorded through 2029.
Conflicting requirement
Termination option effective 2027.
Human review required — assigned to Process owner.
Decision Intelligence
Explain what matters, why it matters and where human judgment is required.
- Findings are synthesised into the effect on the decision, not just stored.
- Source basis, applicable rule and rationale stay attached to the output.
- The reviewer accountable for the judgment is named before anything proceeds.
What matters
Termination option conflicts with the capacity assumption.
Effect on the decision
Committed capacity ends in 2027, so the base case no longer holds on the original assumption.
Required
Accountable for the material judgment. Nothing proceeds until this review is complete.
Synthesised from linked evidence. The model states the effect; the reviewer makes the decision.
Workflow Orchestration
Coordinate models, rules, reviews and approvals in one controlled workflow.
- Intake, validation, model execution and review run as one governed sequence.
- Exceptions interrupt the run instead of passing quietly to the next step.
- No governed output is released while a review gate is still open.
Execution stops at the gate. No governed output is released while a review is open.
Model Runtime
Run governed financial and operational models with reproducible outputs.
- Models run at a known version, against connected and source-linked inputs.
- Outputs are reconciled back to those inputs before they are used.
- The same version, inputs and scenario reproduce the same run.
Version
v4.2
Scenario
Base case
Run
RUN-08
Connected inputs
Key outputs
Same version, same inputs, same scenario — the run reproduces.
Integrations
Connect documents, data and enterprise systems through governed integrations.
- Document libraries, model files and operational systems are connected in scope.
- Access follows the permissions of the connected system, not a second set of rules.
- Outbound actions stay behind an approval boundary.
Connected sources
Access context
Inbound
Outbound
Availability depends on implementation scope — systems are connected with the access each environment permits.
Human Review & Controls
Route exceptions, approvals and overrides to the right people.
- Material exceptions are routed to a named accountable role.
- Approvals, conditions and overrides are recorded with the reviewer's reasoning.
- The material judgment stays with a person, not the model.
Termination option conflicts with the capacity assumption.
Overrides are recorded the same way — with the reason, the reviewer and the accountable owner attached.
Governance & Auditability
Keep every material decision explainable, reproducible and traceable.
- Source reference, model version, rule applied and human decision sit in one record.
- The state before and after the decision is kept, not only the outcome.
- A readable event trail shows who did what, and when.
Record
State change
Before
Capacity assumption unverified
After
Assumption reconciled · condition attached
Audit events
Every material step is reproducible from the record — source, version, rule, reviewer.
Security & Compliance
Protect access, tenant data and governed operations.
- Access is enforced by role, workspace and tenant scope.
- Requests outside that boundary are not served, and the outcome is recorded.
- Architecture and current assurance status are shared during the security review.
Access context
Protected resource
Within role and workspace scope
Outside the permission boundary
Architecture documentation and current assurance status are shared during the security review.
Integrations & Governance
Governed integrations. Governed operations.
Model Runtime executes financial and operational models with reproducible outputs. Governed integrations connect documents, data and systems, with permissioned access, audit trails and human review built in.
Connects with
Financial & operational models
Spreadsheet and structured models, run with reproducible outputs.
Documents & data rooms
Investment memoranda, leases, reports and diligence material.
Microsoft SharePoint
Document libraries and sites connected within implementation scope.
Google Drive
Shared drives and folders connected within implementation scope.
ERP, CRM & operational systems
Enterprise records and workflow systems connected by implementation scope.
Integration availability depends on implementation scope. Enterprise systems of record are connected as part of implementation, using the connectors and access each environment permits.
Governance & Auditability
The Decision Agent
An AI agent that coordinates
the work behind every decision.
It reviews information, works across models and workflows, involves the right people and carries out approved tasks — with every action controlled and recorded.
Connected to
Documents and records reviewed
Assumptions and policies evaluated
Current stage and blockers understood
Tasks routed to authorised colleagues
Approved actions executed through integrations
Instruction
Review the decision inputs, resolve material conflicts and coordinate the required follow-up.
Issue detected
A model assumption conflicts with an active source requirement.
Coordination
Human control
External request awaiting approval
After approval
The agent prepares and executes approved work. The material decision stays with the accountable reviewer.