Venture intelligence, with an operator in the loop, for the people who build and back ventures.
New companies replace what no longer works. The people building them hold one. The system was built for those who hold twenty.
A fund is built to survive any single loss. A founder isn't. But a failure that was visible in time still costs the fund capital and years.
The numbers that end a company, like runway, margin, retention and customer concentration, usually move long before it ends. Someone inside often sees them first. The structure around them rarely acts in time.
Every exit starts with survival. The signs that a company won't survive usually show early, in the numbers, long before the board sees them. No system reads them across the whole company.
Guild Operand is that system. Models drawn from how hundreds of ventures broke, an engine that reads the whole company at once, and operators who make the calls.
We don't write cheques. We make the cheques already written go further.
The Operand Engine™
It starts with the public record and adds founder data when it is shared. Every finding states, in plain words, how confident it is.
Mind, market, mechanics, moat and momentum. Trouble in one rarely stays there, so the engine shows where it spreads. We call this M5.
We keep score. Every read is checked against what actually happened, misses included.
Have a question like these about your company or portfolio?
For the people who build ecosystems
Venture builders, accelerators, VCs and corporate venture arms. A partner sits on more boards than a quarter has room for. A studio runs several ventures with one team. An accelerator carries hundreds of companies and works closely with a few. So few companies get a close read. When one fails, the committee, the LPs or the parent ask what was known and when. The answer is usually spread across decks, updates and partners' heads. The judgment is yours. The record should be easy to produce.
In most portfolios today, each company is read by a different person, at a different time and against a different standard. By the time the committee sees the picture, it has been stitched together after the fact.
Operand Partner is built to change that. It brings one consistent read to every company, so the collisions surface before the board meets and operator attention goes where it is actually needed.
Public record and founder data, for every company, against one standard.
Every part of the business at once, so the collisions show.
For each company, along with which ones need hands this month.
What was seen, when, on what evidence and who decided, ready for the IC, the LPs or an auditor.
Your committee keeps every decision. The engine keeps the record.
Operand Partner is the programme. The first step is a four-week evidence read of a cohort or of three companies, at a fixed fee, written for your committee and LPs.
For founders building to scale
There is one thing you can't do. See the whole company while you run part of it.
The first customers came from somewhere. Whether that somewhere scales is the question nobody on the team has time to ask.
Sales, product and finance each see their own part of why. The collision between them is visible only from outside all three.
Unit economics meet scale. The board sees it a quarter after someone inside saw it and said so.
What goes in
Your public record goes in first. Once we start working together, your internal records go in too.
See every part of the company at once while running one of them.
The engine reads the company across every part at once and shows where the parts collide.
Hold the pattern of hundreds of ventures that broke the same way.
An Operand names the one decision at risk, and what would settle it.
Keep watching the weak joint while shipping.
The engine is then set up for your company and keeps watching that joint.
The decision stays yours.
Capital efficiency is not spending less. It is spending where the company actually holds, and stopping where it breaks.
Operand as a Service is the programme. The engine runs on your company with an Operand beside you, and the decisions stay yours. Bring the company to the call and leave with the one question the engine would ask first.
The first Operand
I have seen ventures from three sides. At two global banks, they were measured as assets. As an operator, I worked inside funded companies. In a venture builder, I co-built a portfolio of 25.
From every seat I saw the same failure. The signs were visible, but no one was placed to see them all. So I built a system that is.
Sandeep Kesavanunni
Founder and CEO, Guild Operand
Thirty minutes with the first Operand. Bring the company, or the portfolio. Leave with the one question the engine would ask first.
Pages for institutions, for founders and for the engine follow on 11 December 2026.