The software company that created FT Mercati Go to ftmercati.com

Method

The FT Support method: from the first conversation to AI in production

Artificial intelligence (AI) in a company does not call for a big project. It calls for a small first step, done well, and the freedom to decide after each step whether to take the next.

  1. First conversation

    About 45 minutes, no obligation

    You tell us how you work and where time is lost. We tell you frankly where artificial intelligence can bring a measurable advantage and where, today, it is not worth it.

    What you get

    • Two or three candidate processes, most worthwhile first
    • A first idea of the timescale and the commitment required
    • No obligation to continue
  2. Process map and priorities

    A day with senior management

    A day with the people who run the company and the heads of department. We look closely at processes and data, collect the proposals of the people who do the work every day and put the processes in order, most worthwhile first. What comes out of it are the decisions that fall to senior management: where AI should deliver a result, in what order, under what rules and with what budget.

    What you get

    • The process map, most worthwhile first, with benefits and difficulties
    • The data inventory: what there is, where it is, how far it can be trusted
    • The immediate improvements, achievable with ready-made tools
    • A proposal for usage rules (AI policy)
    • A step-by-step plan, each step with its cost and expected result
  3. Fixed-scope pilot

    A few weeks, price defined in advance

    A single process, real data, real users. What goes into the pilot is written down; what stays out goes on the list for later. You start from a number and you arrive at a number.

    What you get

    • A working system, used by real people
    • The measure of the result: hours, errors, days
    • What you need to decide whether to go into production
  4. Production and evolution

    Releases in versions, ongoing maintenance

    We add what can be left aside in a trial: role-based access, the action log, error handling, cost monitoring, instructions for users. Then we grow one process at a time, reusing what has already been built.

    What you get

    • Release notes with every version
    • Maintenance, updates and planned new features
    • A second project that costs less than the first

In brief

The FT Support method has four steps: a first conversation with no obligation, a day with senior management on the process map and priorities, a fixed-scope pilot on a single process, and the move into production with maintenance and evolution. Every step has its content, cost and result defined before we start.

  1. 01 First conversation About 45 minutes, no obligation
  2. 02 Process map and priorities A day with senior management
  3. 03 Fixed-scope pilot A few weeks, price defined in advance
  4. 04 Production and evolution Releases in versions, ongoing maintenance

Our rules

Six commitments that apply to every project

  1. Fixed scope, defined price

    No quotes by the hour. Every step has its content, cost and expected result written down before we start.

  2. Every pilot produces a number

    Hours saved, errors avoided, days gained. Without a measurement it is not a project, it is a demo.

  3. People in the loop

    AI proposes, prepares, flags. The person responsible confirms. Irreversible operations stay with people.

  4. Security as a rule

    Credentials never in the code, automatic checks before every deployment, backups before every major change, an action log.

  5. Transparency

    We state where we use AI, including in development. The code is reviewed, tested and put through acceptance testing together with the customer.

  6. Everything documented

    Decisions, versions and instructions live in the project, not in the head of the person who built it.

Frequently asked questions

About the method

Does the first conversation really come with no obligation?

Yes. Its purpose is to find out whether there is a process worth working on. If we do not find one, we say so: that is better for everyone than starting a project without a clear objective.

Can we start directly with the pilot?

Yes, when the process is already clear and the data is accessible. The day on the process map and priorities is recommended when the company has several candidates and no shared priority.

What happens if the pilot does not deliver the expected results?

You have spent little and you have learned where the real problem lies: often in the data or in the process, not in the technology. The fixed scope is there precisely to keep this risk small and known from the start.

Shall we start with the first step?

Forty-five minutes to find out whether there is a process worth working on. If there is not, we will tell you.