Hapax is a small AI consultancy in Helsinki. We call the work advisory, because most of it starts with working out what a problem actually needs before anyone builds anything. Sometimes that’s a Bayesian model built from scratch. Sometimes it’s a dashboard that replaces a week of copying into Excel. If the simple option will do, we’ll tell you, and build that instead.
The work falls into three parts, each with its own page.
Feasibility studies, proof-of-concept builds and deployment.
Closed and in-house, either a standard programme or one designed around your organisation.
A retainer of a few days a month, for clients who want an outside view while a decision is still open.
The scope and fee are agreed in writing before we begin. Whatever we build belongs to you when we leave, and engagements end with a handover so you can run things without us. There are two of us, so we take a small number of engagements at a time.
The usual starting point is a situation, not a specification. The board has asked what the company is doing about AI, or a vendor has pitched something impressive and nobody in the building can judge the claims. Reports or contracts have piled up faster than anyone can read them. A process that one long-serving person understands is about to lose that person, or training was bought and nobody measured whether it changed anything.
One of us spent eight years in financial planning at Procter & Gamble and Coty, running planning for a unit turning over more than a billion euros. The other is a professor of machine learning whose published research covers Bayesian inference and uncertainty quantification. It isn’t a typical combination for an advisory firm, and it’s the reason we take on the full range, from a 40-person company whose data lives in spreadsheets to a problem nobody has modelled before.
There is work we turn down. We don’t build products, supply extra staff, or take on anything that needs a team of ten. A chatbot deployed for its own sake, a project whose data doesn’t yet exist, and any large implementation are better placed elsewhere.
Write to us with the situation, or give us a call. We’re always happy to take a look, and will gladly offer advice whether or not we can take on the work.
ἅπαξ λεγόμενον. The thing said once.
A hapax legomenon is a word that occurs only once in a body of text. With no second instance to compare it against, its meaning has to be reasoned out from the context around it.
The problems that reach an outside adviser usually have the same structure. Each arrives in one specific configuration, with nothing closely comparable to check it against, so the answer has to be worked out from the evidence available.
Associate Professor of Artificial and Human Intelligence at the University of Helsinki, where he leads the Machine and Human Intelligence group.1 He is also Associated Faculty at the ELLIS Institute Finland.2 His doctorate in computational neuroscience and machine learning is from the University of Edinburgh.
He has published more than thirty-five peer-reviewed papers, over fifteen of them at NeurIPS, ICLR and ICML, with others in Nature and Nature Neuroscience,3 and wrote BADS and VBMC, open-source inference software used in research groups worldwide.4 His papers have been cited more than two thousand times.3 His research area is uncertainty quantification, which in practical terms means knowing how far a model’s output should be trusted. At Hapax he is responsible for technical design, quality assurance and the advanced teaching.
Runs the firm. She leads the advisory engagements and delivers the training, teaching the advanced sessions alongside Luigi. She has an MSc in Data Science and AI from Goldsmiths, University of London, a deliberate step out of pure finance to understand what the technology can actually do.
Before that she spent eight years in financial planning and analysis at Procter & Gamble and Coty, latterly responsible for planning at the $1.3B Wella Professionals unit. That meant distributor analytics covering more than forty markets, a $2B division’s forecasting through a divestiture, and a replacement forecasting system in Oracle built from scratch. She was also Coty’s data-analytics trainer for Europe, designing and delivering the programme across the region, work close to the training Hapax does now.
Three commitments shape every recommendation we make.
Some of what we learn gets written up. At the moment we are working on a paper on when to trust a language model’s output, which will appear when it is finished. Eleven case studies are published, each built on data where the correct answers were written first, so the figures in them can be checked.
Tell us the problem, what is at stake, and when you need to move. One of the two of us will reply within two days.