We are Pareto AI

Welcome to the Pareto AI blog! We are a team of artificial intelligence experts who work with other businesses to help them on their journey in the AI world. We spend all our time working on cool products, but rarely take “me” time as a team to share what we are building and creating along the way. And since we’re often asked about our projects, learnings, ways of working and of course views on AI, we decided to start a blog to share all of that and much more…

How it all started

Pareto AI was founded by two AI enthusiasts, once PhD colleagues, who have been in the AI industry for almost 15 years and have many real world AI products under their belts. Dejan was a Director of Data at Gousto, a meal delivery startup based in London that got a unicorn status with SoftBank investment. Miha was a Chief Data Scientist at Golden Set Analytics, a startup specialised in tennis analytics, working with top tennis players in the world.

Back in 2020, after Dejan’s “Gexit”, they decided to continue doing what they do the best, i.e. building AI products. As they wanted to stay close to AI, but also to continue enjoying an excitement of diverse problems, they started a business to help other businesses on their artificial intelligence journey.

Pareto AI was born. Coming from the industry, it was clear that our motto should be: “we build AI products from zero to production, with an emphasis on business value. ”, which we learnt is the most challenging part of artificial intelligence projects in the real world.

What we actually do

As we are partnering up with ambitious clients to build AI products, we bet on focus and transparency. We understand that we don’t live in a perfect world, meaning some things don’t work as planned, especially in the artificial intelligence world, where the output heavily depends on the data – and we all know that the data can be messy and is changing over time.

To make progress as smooth as possible, we work in agile sprints, delivering value incrementally and iteratively. Such an approach ensures that we are all on the same page all the time and the right priorities are taken into account as we go.

Following the same idea, we communicate frequently and transparently with our clients. We proudly share great results as part of regular reports and demos, but we are also not shy to share our failures as we believe that learning is part of growing and transparency is key to strong relationships.

Our definition of done is not based on delivered code or models, but rather on delivered solutions that solve real problems and create real value. We care about our clients’ success as much as ours.

Our story so far

Since our inception three years ago, we have grown organically to a team of almost 20 artificial intelligence experts, working with various clients. With some of the clients we’ve been working for more than 2 years now, tackling challenging problems, which we really enjoy and give us strong motivation.

Although we find excitement in almost any problem that can be solved by artificial intelligence, with our experience we’ve specialised in the following areas:

– time series forecasting, e.g. demand forecasting, energy consumption/production forecasting, churn prediction,

– optimisation e.g. marketing channel mix optimisation, inventory optimisation, pricing optimisation and

– text processing, e.g. ChatGPT for customer care, chat assistants on internal documents, information extraction from free text or PDFs, news articles generation etc.

As we continue growing and working on cool projects, we will be publishing more about our work, learnings and thoughts in the upcoming blog posts and case studies. For starters you can read our most recent publications:

Extracting information from price lists using AI in which it’s described how we helped to build AI tool to extract key information from price lists for market analysis and sales optimisation,
Helping journalists to focus on creating premium content that outlines our approach to automate the most tedious tasks in journalists’ workday with a focus on preparing news articles that summarise data-rich events, e.g. sports news such as NBA games etc.

We look forward to the new challenges, not only associated with our projects, but also related to our growth and learnings on our journey to become the go-to team when it comes to artificial intelligence solutions.

Last but not least

Have you been wondering why we named our startup Pareto AI? Time is of an essence in the business world and delivering value as early as possible is crucial. With that in mind, we are big fans of the Pareto principle also known as 80-20 rule, which states that 80% of the results come from 20% of the efforts. We apply this principle to our work, focusing on the most important features and tasks that can generate the most value for our clients. Hence the name Pareto AI.