Blinko

How sure are you?

A modern statistics and AI boutique.

Before you change the price, approve the loans or start the trial, we show you what the numbers really say, what could go wrong and what to do next.

See the examples
A live simulation. As customer data arrives, a statistical estimate of a rollout's effect narrows from wide and unsure to sharp, moving to one side of break-even until the evidence supports a decision.

Clients include

Fractal

Solventis

OmniXM

Apickle

You bring the question. We tell you what your numbers can answer, and what they cannot.

  • Sales rose after the promotion. You see how much you caused, and how much was just spring.

    The changeWhat happenedReal effectWhat would have happened anyway
    E[ Y(1) − Y(0) ]See how we do it

The hard problems are the ones we take. Selected engagements.

Fractal and Sky Systemz

Health scores, forecasting and an analytics system with AI, running the business every day.

A ranked list of accounts by risk, with the signals that moved one account's score.
Illustration
$100M+
in yearly transactions tracked
“Blinko delivered exceptional work on a highly complex problem. Deep expertise in data science, machine learning, and business domain knowledge – exceeded all expectations.”
Steve Conn, Analytics Director, Fractal & Sky Systemz

Noise has rules.

Each ball falls at random. Together they draw the same curve every time. Know what chance looks like, and you can spot what is not chance.

Nearly 100
client engagements delivered since 2023
2 papers
published, in embryo assessment and retinal genetics
Georgia Tech
MSc in Computational Statistics and Analytics
Bank of Israel
statistical and risk analysis

  • Finance runs on risk. We put a number on it.

    When the board asks how sure you are, you have a number you can defend. Credit, fraud, customer loss and revenue risk.

    • Who will default, and what would it cost us?
    • Is this fraud, or a normal bad day?
    • Which merchants are quietly leaving?
    Statistics for finance

  • Healthcare needs evidence that survives peer review.

    When a journal or an investor asks how sure you are, your evidence holds up. Study design, patient outcomes and testing of clinical AI.

    • How many patients does this study need?
    • How should the trial be designed?
    • How long until the event, and what changes it?
    Statistics for healthcare

Published research

Embryologist agreement when assessing blastocyst implantation probability: is data-driven prediction the solution to embryo assessment subjectivity?

First author. Oxford University Press, on behalf of the European Society of Human Reproduction and Embryology.

Published research

Mathematical Model to Assess Mutational Burden in Retinal Dystrophy Patients Negative to Mendelian Genetic Tests and Carriers of Multiple Recessive Variants

Co-author. Journal of Ophthalmic and Vision Research. A preliminary model, studied in 527 subjects.

All case studies and notes

Bring us the hard question.

Tell us the decision you have to make. A statistician will tell you, honestly, whether your data can answer it.

NDA first. Read-only access or a secure export. We keep no copy of your data after the project.

  • Start with a diagnostic

    In two weeks on your own data, you know what is driving the problem, what it is worth and what to do first.

    How we work

  • Need the AI side too?

    Assistants, agents and live analytics systems, built and tested to the same standard.

    AI solutions