Health monitoring for private-company investments
Know when a company you’ve backed needs a decision, not just how it’s doing.
Pre-seed · Building with design partners · Raising our first round
The problem
An investor commits capital, the round closes, and the company goes quiet. What comes back is a deck before a board meeting, written by the people being assessed, in a format that is different for every company in the portfolio.
The numbers that matter arrive late, and they arrive smoothed. A runway that was tight three months ago is a crisis today. A key account that stopped renewing in spring shows up in an annual summary.
By then the window has narrowed or closed for the choices that were open: a bridge, a pivot conversation, a new round, an early exit. A turning point you notice only in hindsight is no longer a decision.
How this is handled today, and what we do differently
Risk assessment systems already exist, and they work well where they were built to work: on companies whose numbers are audited, whose filings are public, and whose failures are recorded. Early-stage private companies are the opposite case: self-reported numbers, quiet failures, and little historical data to learn from.
Investors have covered that gap the way good investors always have: board seats, direct relationships with founders, and judgement built over many deals. It works, and it does not scale. Ten companies, then thirty, and the quiet ones drift out of view.
Carota is built to hold that attention steady across the whole portfolio, without replacing the judgement that goes into using it:
A model that learns. We train it on a growing base of company histories to recognise the patterns that come before trouble, and the ones that come before a breakout.
Continuous observation. Each company stays in view between board meetings, not only when a report is due.
A signal, not a stream. The system sends a signal when a company reaches a turning point, while the window for a decision is still open.
Light on founders. A handful of plain numbers each quarter. The interpretation is our job, not theirs.
Data carries its origin. Numbers are marked by how they were confirmed, inconsistencies surface instead of being smoothed over, and history is kept rather than overwritten.
Explainable, factor by factor. You see what drove an assessment, not just a score. The model advises. The investor decides.
Isolated by customer. Each customer’s data is kept separate and used only with the portfolio company’s consent.
Who it’s for
Carota is built for investors holding equity in private companies.
Angel investors
A structured view of a portfolio that is otherwise tracked in a spreadsheet and a memory.
Venture funds
Comparable companies, and traceable assessments for partners and LPs.
Accelerators & incubators
A whole cohort measured by the same yardstick.
Corporate venture teams
Visibility across investments made outside the core business.
This is about investments in private companies. Portfolios of listed securities are a different problem, and not one we work on.
Founders get something out of it too: a fair, structured assessment of their company in return for a few minutes of reporting a quarter, plus a record of what they lived through, kept intact for the next round.
Where we are
Pre-seed. The methodology is in final development. The scoring model is built and trained on an initial set of company histories; large-scale training is the next step, and the work ahead is to sharpen the timing of turning points. The platform itself is in development.
We are building with design partners: angels, funds and accelerators who run their portfolio through Carota while it is being developed and shape it with us.
Team
CFO of Ovoca Bio plc (LSE/ISE) for a decade: 20–30 project assessments a year, M&A, a fundamental business pivot and re-listing. Earlier, asset sales to international buyers in a holding of 70+ companies. Now an independent advisor in financial modeling, valuation and corporate governance. PhD in Economics, ACCA.
LinkedIn →
Software engineer with a career in companies that won their markets: Silverbyte (acquired by Priority Software) and Perimeter 81 (acquired by Check Point). Today builds production ML underwriting systems at a US insurtech. Financial platforms, rating engines, explainable models, data pipelines.
LinkedIn →Talk to us
Svetlana Radchenko, CEO
svetlana.r@carota.vc