Our work

A selection of recent projects, described without any confidential data. We can go into more detail on some of them when we talk.

AI in production

The need: IT support received a large number of simple, repetitive requests.
What we did: A chat assistant built into Teams answers users' questions from the internal documentation, and automatically creates a ticket in the tracking tool when a human needs to step in.
The result: Users get an immediate answer to common questions, and the support team focuses on real incidents, with tickets that arrive already qualified.

The need: After every call, customer service advisers spent time writing up a report.
What we did: A language model analyses the call transcript as the conversation unfolds and suggests a summary directly in the CRM.
The result: Less typing for advisers, and consistent call reports available as soon as the call ends.

The need: A shared inbox received a large volume of messages that had to be routed by hand.
What we did: An AI tool reads each incoming email, works out what kind of request it is and routes it to the right team.
The result: Requests handled faster, and less time spent sorting.

The need: Connected products had to communicate reliably with web and mobile applications and with the information system.
What we did: Technical lead for the connectivity backend and the APIs linking the applications to the information system, in a microservices architecture on Azure.
The result: A shared platform, reused by several applications and product ranges.

The need: Each application managed its own accounts, which meant more and more passwords and made security harder.
What we did: Implementation of a cross-cutting identity and API management solution (SSO), used by customer-facing and employee-facing applications alike.
The result: One account per user, and centralised access management.

The need: An application infrastructure that needed modernising and making easier to run.
What we did: Contribution to the architecture choices and the migration to Azure, then administration of the resources: environments, costs, security and automated deployments.
The result: Automated deployments and an infrastructure under control.

The need: The same customer's details could differ between their online account, the CRM and the ERP, leading to errors and data being keyed in twice.
What we did: Development of a back-office tool that compares the data from the different sources, highlights discrepancies, and lets staff merge or correct accounts in line with the rest of the information system.
The result: Customer data that matches from one tool to the next, and corrections made in a few clicks rather than by hand in each system.

The need: Activity tracking relied on files consolidated by hand.
What we did: Development of internal tools that collect the data and calculate the production KPIs automatically.
The result: Reliable, up-to-date KPIs, with no re-keying.

The need: A company wanted to replace its spreadsheets with a management application suited to the way it works.
What we did: As Product Owner: defining requirements with the teams, writing and prioritising the backlog, following development and running acceptance testing.
The result: An application built around how the teams actually work.

A no-obligation first conversation, to understand what you need and see whether we can help.