In practice

Situations you may recognize

These projects are real. Out of discretion, the companies are not named: their story is theirs to tell. Here are the starting problems, the choices made along the way, and what changed day to day.

Business services

One list of offers — and data people can rely on

The starting point: The offers users cared about came from a long list of scattered sources. Everyone ran the same searches on their own. Nobody could say whether a list was complete, or whether the information was still right.
What we did: A tool collects those sources on a regular schedule and puts every offer into the same format. A search engine then filters them on the criteria that matter to the people searching. One deliberate choice: nothing is published automatically, and a doubtful source is set aside rather than risk showing bad data.
What changes: One list, kept up to date on its own, instead of the same searches repeated source by source. And information that holds up: what appears has been checked, what is doubtful does not appear at all.

Industry

Field readings come back on their own — and never leave the company

The starting point: Technical readings taken out in the field, then re-typed on return — with the losses and errors that come with it. And handing that data to an online service was out of the question: it had to stay in-house.
What we did: A mobile app records each reading right where it is taken. The data flows to a server that files it as it arrives, and dashboards make it readable straight away. Everything was installed on the company’s own servers: nothing goes to the cloud (servers rented from a third party). That choice means more setup work up front — and it was made deliberately.
What changes: Readings arrive once, clean, and everyone looks at the same figures. The data never leaves the company.

Services

One place to drive all the tools

The starting point: The working day meant juggling several applications: open one to check, another to type, a third to send. Every switch cost attention, and things got lost along the way.
What we did: An internal assistant was connected to the tools already in place. The team speaks to it, or types, in plain language, and it acts: find a piece of information, create a task, prepare a message. It runs locally (on the company’s own machines), with no outside service involved. The existing tools were kept, not replaced — that was the condition for the team to adopt it.
What changes: One front door instead of a row of windows open all day. Everyday requests get handled where people already are, without switching screens or retyping.

Public procurement

Choosing which tenders to bid for, instead of hunting for them

The starting point: For this team, the time that counts goes into deciding whether a contract is worth pursuing, then building the bid. It was going elsewhere: every morning, someone combed through the same platforms, with the nagging fear of missing the one that mattered.
What we did: A tool now watches the relevant sources continuously and filters notices against criteria set with the team: contract type, area, keywords. Only relevant notices trigger an alert. The filter was deliberately set wide at first, then tightened with use — better to sort a little too much than to miss a deal.
What changes: The morning review is gone. The notices that concern the team arrive on their own, and its mornings go to the only question that matters: do we bid, or not?

Training

Training that works anywhere, even with no connection

The starting point: A training provider needed a tool, not a trainer. Its sessions happen in rooms where the connection is poor — or deliberately cut. Online materials failed at the worst possible moment, in front of the audience.
What we did: The full course was built to live on the computer itself: content, exercises, progress tracking. Nothing leaves the machine, and no data is sent anywhere. The trade-off was clear: give up the features that need the internet, so that everything works, every time.
What changes: The session runs the same with or without a network. The trainer walks into the room without wondering whether it will work, and participants’ data stays where it is.

These projects are shown without names or precise sectors: our clients have not given us permission to name them, and we would rather publish nothing than do so without their consent. We quote no performance figures either: we did not measure these projects, and we will not invent results.

So where do you start?

Describe your situation: you get a heading, a scope and a quote in return. No commitment.

Build my quote Describe my project →

Reply within 48 business hours · No commitment · Confidential