The AI Models Behind StructureFlow and Why They Matter

The most complex work in professional services doesn’t arrive in a clean format. It comes as layered ownership charts buried in PDFs, transaction chronologies scattered across documents, fund structures built across jurisdictions, deal histories that span years of emails and filings. The challenge has never been finding the information. It’s been making sense of all the input into a deal or matter fast enough to gain competitive advantage. Complexity in how deals or matters evolve or change is increasing and putting more strain on everyone evolved.  

This is the problem StructureFlow was built to solve. As AI evolves to help empower how we deal with complexity in our working worlds, it’s a natural evolution for StructureFlow to start to take advantage of the tools available and stay ahead of the curve by including the opportunity for our clients to leverage the leading AI LLMs. 

The platform for complex visualisation 

StructureFlow is built for a specific and demanding kind of work: making structural complexity visible, navigable, and actionable. The firms that use it, M&A lawyers, tax advisors, PE funds, restructuring specialists, deal with some of the most intricate relational data in any industry. Ownership chains, counterparty networks, deal structures, litigation chronologies – the kind of complexity that defeats spreadsheets and overwhelms slide decks. 

Turning that complexity into clarity requires more than good software. Turning that complexity into clarity requires AI tools that can read an unstructured document and build a structure from it. One that answers the question “who ultimately controls this entity?” across a chain of indirect ownership and shows the working. 

That’s structural intelligence. And at StructureFlow, we continue to make it significantly more powerful with every product update. 

You now have a choice. Why we’ve added the latest AI models to StructureFlow 

Two things drive this decision. 

The first is capability. The work our clients do rarely arrives in a single clean format, it comes as PDFs, images, Word documents, emails, and spreadsheets, all of which need to be read, interpreted, and converted into something workable. By incorporating new AI models we can enhance performance across all of these: stronger image recognition, more accurate document processing, and significantly higher context limits.  And beyond individual tasks, there’s a broader principle: we don’t want StructureFlow to be constrained by what a single provider’s models can do. The frontier of AI moves fast. The firms we serve need tools that keep pace with it. 

The second is resilience. StructureFlow is used in live deal environments, under time pressure, where things going wrong is not an option. Relying on a single AI provider creates a single point of failure. By supporting multiple providers like OpenAI, Anthropic, Google, we’ve ensured that if one has an outage, the work doesn’t stop. You can switch models and carry on. For organisations where continuity matters, that’s not a minor point. 

What is “better AI” and what does it actually enable? 

The AI Assistant in StructureFlow lets you work with a structure the way you’d work with a knowledgeable colleague. Ask who the ultimate beneficial owners are across a chain of indirect holdings. Ask what happens to downstream entities if a parent company sells its stake. Ask the platform to insert a new holding company into an existing structure and reroute the ownership lines. Ask it to flag every entity incorporated in a specific jurisdiction. 

With the new models all these interactions become even faster, even more accurate, and capable of handling structures of a complexity that would have strained previous context limits. Lawyers can interrogate deal structures in the room, in front of clients. Tax advisors can model reorganisations on the fly. Fund managers can stress-test ownership positions without leaving the canvas. 

This is what structural intelligence looks like in practice. Not AI as a party trick, but AI as a genuine co-pilot for the most demanding work in professional services. 

Built for businesses in regulated environments 

The firms StructureFlow serves operate in regulated environments. They work with sensitive client data. New capabilities don’t get adopted because they’re impressive, they get adopted because they’ve been evaluated, understood and signed off. 

The new models are off by default.  

We know control for client is important so enabling them is a decision for firm administrators, not individual users. The documentation, FAQs, data processing details, and security information is all available to the compliance and IT teams who need to work through it properly. We’d rather organisations take the time to make the right decision than rush into something they haven’t had a chance to assess. 

The capability is there when they’re ready.  

The platform is already one of the most advanced tools available for structural work in professional services. The AI powering expands the opportunity to work smarter, not harder, and ultimately, win the room.  

Interested to see this in action? Book a demo today.