Meet Iggy.
A voice and text agent that queries the IGX360 canonical model directly. It explains what you're looking at, navigates you into the right in-app view, and connects findings across all five lenses. The experiential proof of Stage 3.
Watch Iggy work.
A single impact-analysis question, asked in plain language, answered against the canonical model with traceable provenance — then relayed live in French, German and Japanese for colleagues in the room, without losing context.
Follow the language switches
Iggy answers the same question six times over, in four languages, without re-running the analysis. Click any moment below to jump straight to it in the video above.
Iggy finds the KYC tool application and runs a full impact analysis across processes, capabilities and business services.
Answers "who needs to be informed" and opens the graph and depth view to show the full notification scope.
Relays the same graph-view summary and notification answer live in French, for a French-speaking colleague.
Relays the same summary again in German, without re-running the analysis or losing context.
Explains the KYC checks and screening steps in Japanese for visiting partners.
Switches back to English to close out the session.
From a single question, to a defensible answer.
Select any panel to open the full-bleed carousel. Read the moments in order, or skip to the reasoning step.
Conversational where it helps. Defensible where it matters.
Iggy is built for enterprise BPM: auditable, transparent, and grounded in the canonical model. No hallucinated answers.
Canonical SQL
Every answer is traceable to a SQL query against the IGX360 canonical model. No retrieval black-box. The query is shown, audit-ready.
Confidence scoring
Each response carries a calibrated confidence figure. Below threshold and Iggy declines to answer, prompting you for clarification.
Voice and text turn-take
Talk to Iggy, type to Iggy, or both in the same conversation. Each modality preserves the same audit trail and citations.
Prompt injection resistance
Domain-bounded retrieval, canonical query whitelisting, and adversarial-input filtering. Iggy answers from the model, not from the prompt.
Frequently asked questions
What is Iggy?
Iggy is the voice-first conversational interface for IGX360. It translates plain-language questions into canonical SQL queries against the IGX360 canonical model, returning answers with provenance: the source object, relationship path, and confidence score for every response.
What data does Iggy query?
Iggy queries the IGX360 canonical model only: the reconciled dataset built from your iGrafx repository, Visio diagrams, SOPs, and BPMN/XML files. It does not access external data or the public internet.
Does Iggy use my data to train AI models?
No. IGX360 operates on a BYOK (Bring Your Own Key) model. Your organisation provides its own LLM API key, credentials are held in session memory only, and are never transmitted to IGX Solutions infrastructure.
Bring Iggy a question. See what your repository has been hiding.
A 15-minute live session against a sample model. We'll show you the gap-revelation moment with your kind of repository, then leave you a written read-out.