What we do
An Insurer, a MGA, and an Investor walk into a bar...
...and live happily ever after.
Data Efficiency & Analytics
Pricing Algorithms & Framework
Portfolio & Capital Structuring
Data and Growth doesn't have to play chicken and egg. With the latest techniques and domain expertise, we can squeeze more out of what you have already got.
Only a few quarters' worth of data in a new line of business? Not a problem.
Drowning in data and need to prioritise the needle-moving markers? We can help.
Don't want to miss out on AI, but loathe to share the same model as everyone else?
Let us fine-tune it to your field-specific mastery.
Your expertise is valid. Domain knowledge built-up over decades should not be dismissed out of hand.
We translate your battle-tested "norms" and logic into quantitative forms, and entwine it with the observed data using explainable models.
It is never an insight if it's simply spit out from a black-box. This is why our algorithms are designed to be transparent, and are here to enhance your decision process, not to leave it out.
Want to build a new rating model from ground up, or get our take on the complete pricing journey from data capture to quote? You got it too.
Time is money. Why wait until it's too late to take action?
Be it individual binders, a portfolio of programmes, or an international multi-line book, stakeholders need to balance risk and return.
It is important to identify the expected performance in each - quickly.
But to understand how likely it is for it to go pear-shaped or stratospherically awesome? Or how to structure it so it's within shareholders' appetite? Come talk to us.
Indistinguishably Magic.
Next-gen open-source language
On top of R & Python, we embrace Julia - taking advantage of a language designed to walk like Python but run like C.
Built for speed and massive parallelism, it has also been increasingly adopted by academia - shaping a new generation of cutting-edge algorithms and libraries for the real world.
Bayesian approach
Three things are certain in life: death, taxes, and uncertainties.
A bayesian approach explicitly measures the latter, helping you take smarter first steps towards balancing risks and rewards.
We regret to say we can't help with the first two.
Explainable Neural Networks
Black-boxes are not insights. This is why, despite being increasingly outperformed, Generalised Linear Models (GLMs) are still prevalent in the insurance industry.
With recent advances however, we can now conveniently approximate and recover analytical equations (like those in the GLMs) from these AI/ Machine Learning systems.
This lets us showcase the best of both worlds: human comprehension and machine predictiveness. Glue these with the power of the Julia ecosystem and the Bayesian approach - that's when the show begins.
Fine-tuning Large Language Models
Large Language Models (LLMs) like OpenAI's ChatGPT are powerful because they have been trained from a humungous amount of General Information.
But what you do is specialised.
A fine-tuned model re-wires some of those parameters to adapt it for particular domains, making it speak your language, improving its accuracy and effectiveness.
