Reconciling a sponsor's term grid shouldn't take a weekend
Upload a sponsor's term grid in any format. Tusk Grid reads it, matches it against your firm's own closed deals, and returns one standardised, comparable grid.
No you are not imagining it. Sponsor Grids have become super complex.
Not long ago, a sponsor grid was a formality. A vice president received a template, filled in the blanks, and returned it so the real work of negotiating a credit agreement could begin. However, in the past several years, sponsors have transformed their grids into dense, highly specific negotiation instruments, each one calibrated to sector and sometimes to individual deal. The same sponsor will send a materially different grid for a healthcare platform acquisition than for a software roll-up. Grids have become so complex that some sponsors have handed the process to outside law firms entirely, who manage successive rounds on the sponsor's behalf. The negotiation, in other words, now starts before anyone acknowledges a negotiation is underway.
If that shift were merely aesthetic, it would be manageable. But sponsors have added another dimension: time pressure. Grids increasingly arrive with explicit deadlines for response, compressing the window in which a deal team can do the analytical work required to fill them competently. The combination of greater complexity and shorter fuses has made the grid a source of genuine operational stress for lending teams. A VP whose inbox has grown heavier over the past two years is not imagining things.
The Archaeology of a Single Cell
What makes the sponsor grid so difficult is not the grid itself but everything that must happen before a single cell can be filled. The first task is finding comparables (aka comps), which turns out to be a form of institutional archaeology. The relevant deals may have closed years ago, on different desks, in different geographies, sometimes under different group structures within the same firm. Synthesizing these transactions quickly enough to meet a sponsor's timeline is, for many teams, a harder problem than the grid ever was.
Once a few comps surface, the real labor begins. Each one requires opening an entire deal folder and navigating credit agreements that run to hundreds of pages, searching for the precise data point that answers one field in the grid. The analyst finds it, records it, and moves to the next cell. Then the next comparable, and the same process again. It is painstaking, repetitive, and error-prone in exactly the way that long, detail-intensive manual work tends to be.
And the process does not end when the grid goes back to the sponsor. What returns is a counter-grid: a new iteration built upon the lender's response, with revised terms, additional fields, and occasionally a restructured layout. Version control becomes its own problem. Deal teams find themselves checking and rechecking whether they are responding to the most current version, because responding to the wrong one is not a theoretical risk. It happens regularly, a predictable consequence of tracking intricate data across multiple documents under time pressure.
Why Automation Kept Failing - Until now.
The obvious response to all of this is to automate the grid. The less obvious fact is that previous attempts have run into a set of interlocking problems, each difficult on its own and collectively resistant to partial solutions.
The first and most stubborn is accuracy. In grid work, accuracy is not a feature; it is the entire point. Every cell in a sponsor grid carries weight. Deal terms, pricing, covenant thresholds: the content of a single field can shape the trajectory of a negotiation. Any system that cannot deliver auditable, verifiable results at every step is not saving time. It is manufacturing risk.
The second problem is translation. Because sponsor grids change from deal to deal, an automated system must understand what each cell is asking and map it to the correct concept within the firm's own taxonomy. Even assuming such a taxonomy exists (and many firms have only informal ones), the mapping is rarely one-to-one. The same term can mean different things in different contexts, and placing content in the wrong conceptual slot defeats the purpose.
The third is format. Grids arrive as PDFs, Excel files, and Word documents, sometimes all three across a single deal's lifecycle. A system that works only with one format forces someone on the deal team to spend time converting the others manually. Anyone who has copied a structured PDF into a spreadsheet cell by cell knows how joyless and error-prone that conversion is.
The fourth problem is the comp search itself. A firm that has invested in hundreds of deals over many years does not always have a structured way to explore them. Before comparables can be retrieved, the underlying technology must read deal documents, decompose them into discrete attributes, and make those attributes searchable in a way that is fast and intuitive. Without that foundation, finding the right comp remains a manual expedition through filing cabinets that happen to be digital.
The fifth is versioning. Filling the inbound grid accurately is only the first move. The sponsor's counter-grid introduces new fields, comments on existing ones, and builds upon the lender's response in ways that are easy to lose track of. Keeping every iteration visible in a single pane, so that no one responds to the wrong version, sounds simple. In practice, the volume of data and the visual noise surrounding it make mistakes almost inevitable.
TuskGrid: Full Automation, Full Control
TuskGrid, built on CredCore's credit-native AI, addresses these problems as a system rather than piecemeal. Lenders upload the grid as it arrives, in whatever format it happens to be. The platform abstracts away the difference between PDF, Excel, and Word, so the deal team never has to convert anything manually.
From there, lenders use the system to identify comparables in minutes rather than days, directing TuskGrid to extract the relevant data from those deals and populate the grid's fields. CredCore does this at a level of accuracy consistent with outside counsel, at a pace that was not previously possible.
Once the grid is filled, it is ready for the lender to download in Excel. The lender decides what to share and when. No communication flows directly from the platform to the sponsor. The process stops at the lender's discretion, preserving the control that deal teams rightly insist on.
TuskGrid tracks every iteration across the deal's full timeline. A new version of the grid, one where the sponsor has added rows or revised terms, can be uploaded and compared against prior rounds immediately. Every result the platform provides is auditable and editable. The system does the work. The lender makes the decisions.
We built Tusk Grid because that shouldn’t be this much of a grind to battle with grids.
Saumil Annegiri, co-CEO