Projects

We design and build software

a company works on every day

The problem has to be understood, an architecture designed, code written that holds up, and someone has to follow it all the way to production. We have been doing this for years, on platforms, applications and data systems. AI comes in where it is needed, with a defined task and a way to check what it did.

The software we have always built

Enterprise architectures

Integrated, secure platforms built to withstand growth.

Learn more

Web and mobile development

Custom applications, native and cross-platform.

Learn more

Data

Pipelines, analysis and large volumes. On ClickHouse we work with certified consultants.

Learn more

Cloud

Cloud-native systems and migrations, on AWS and the other providers. We keep an eye on what they cost, too.

Learn more

And the four AI projects we are asked for most

Assistants and agents on your data

They answer questions, prepare documents and carry out tasks inside your systems. They cite their sources and respect the permissions of whoever uses them.

Document reading

Contracts, financial statements, forms and e-mails with attachments become structured data. Every field has a confidence level, and when it is low a person checks.

Predictive models

Deep learning, time series and data correlation to forecast failures, consumption and demand. Built on your data.

AI inside the platforms you already have

A new feature in an existing product, without having to rebuild it.

Design, development, operation

  1. Design

    We start from the problem and the numbers. What the system has to do, how you tell whether it does it well, what each answer can cost. The model is chosen afterwards.

  2. Development

    Our engineers write the software and leave the mechanical work to the agents. We measure the quality of the answers for the whole length of the project.

  3. Project management

    One of our managers keeps track of time, cost and risk, and tells you every week where things stand.

Project by project, one step at a time

We start with a two or three week pilot build, which is a piece of work in its own right and leaves you a result even if you decide to stop. Then the full build, and if needed ongoing care for a monthly fee. We write down the duration and price of each phase before we begin.

Where we have done it

A few jobs, by sector

We do not name clients, for confidentiality. The sector and the shape of the work we do.

Specialty insurance

Tailor-made policies underwritten with AI

A specialty insurer built its own underwriting platform with agent coding: the customer describes the risk, the AI composes the tailored policy and prices it, and the underwriter steps in only on risks outside appetite.

  1. Risk intake from questionnaire, documents and external sources
  2. Clause composition and pricing by AI agents, within written appetite limits
  3. Automatic issue; risks over threshold go to the underwriter with the reasoning already written

The underwriter moves from filling in policies to deciding on the risks that matter. The platform was written entirely with coding agents, measured like everything else.

Retail banking

Account opening and KYC without retyping

ID document, proof of address and questionnaire arrive through different channels and end up in three systems. The case assembles itself, and the operator only sees the ones with a doubt.

  1. Reading of documents and selfie, with consistency checks across sources
  2. AML and watch-list checks by rules, and an agent for the name-match cases
  3. Opening in the core systems via API, or via RPA where there is no API

On clean cases opening goes from days to minutes; the operator works the exception queue, not the whole queue.

Business lending

Credit files ready before the committee

Financial statements, credit bureau data, company records and business plans read and reconciled on their own. The analyst receives the file already worked up, with the ratios computed and the points to probe highlighted.

  1. Structured extraction of statements and records, reconciled across sources
  2. Ratio computation and comparison against the credit policy
  3. Pre-assessment written by the agent, with sources cited field by field

The analyst starts from judgement instead of data collection. Every number in the file has its source, and the committee sees it.

Questions we get

What people ask before signing

The one that fits the job. A small, fast one to classify, a more capable one to reason, a local one when the data cannot leave. Applications talk to a gateway, so changing model becomes a configuration change.

Contact us

A quote, a particular request, or simply a coffee to get to know each other? Write to us and we will reply in a "nanosecond".

Address
Piazza Maestri del Lavoro 7
20063, Cernusco sul Naviglio (MI)
Italy
Address
Piazza dei Martiri 1
40121, Bologna (BO)
Italy