Intelix

How we work

Our method comes from running data and AI programmes inside investment firms and the companies they own. It is built for owners and management teams who need results they can see in the accounts.

  1. 01

    We start from the value case

    Before any discussion of tools, we ask what value means for this business. For an owned company that is usually the plan the board has signed off and the investment case behind it. For an investment firm it is the time and quality of its own decisions.

    We then build a value tree. It links every proposed initiative to a line that moves: revenue, gross margin, operating cost, cash or capacity. An initiative that cannot be placed on the tree does not start.

  2. 02

    We choose work with three questions

    Every candidate initiative is scored on three questions, and the portfolio is built from the ones that score well on all three.

    • How large is the prize?

      The net economic gain after the cost of building and running it.

    • How soon does the money arrive?

      We count the time until value shows up in the accounts. A tool built in six weeks that takes a year to change behaviour is a slow initiative.

    • How likely is it to work?

      Whether the data exists, the technology can do the job, and the people who must change are ready to.

  3. 03

    We run two tracks in parallel

    The first track delivers value: proofs of value, then scale-up of the ones that work. The second track builds foundations: data, platforms, access and governance.

    The foundations track is paced by the value track. We build the data and platform pieces that the next initiatives in the queue need, in the order they need them. Governance stays light, and business and technology owners write it together.

  4. 04

    We start small and prove it

    The first team on any problem is small and senior, and works on one well-defined question with full backing from leadership. One category, one region or one process. Once it proves the result, we scale it. This builds evidence and support inside the organisation before large budgets move.

  5. 05

    A sprint, stage by stage

    Most transformation work runs in sprints, one function at a time. A typical sprint has four stages.

    1. Stage 1

      Baseline

      We agree the function, measure how it performs today, check data and system access, and set the criteria a proof of value must meet.

    2. Stage 2

      Future state

      The function head and a few of their people design how the function should work with AI inside it, starting from the target and working back. We then pick three to five proofs of value.

    3. Stage 3

      Build and test

      Each proof of value runs on real work with real users, built and changed in short cycles. We expect some to fail, and we stop those quickly.

    4. Stage 4

      Close

      Results are measured against the baseline. What worked is written into budgets, service levels or the organisation chart, and the playbook passes to the client team.

    Several functions can run sprints in parallel. Investment-side functions tend to move faster. Back-office functions often need more time for data.

  6. 06

    We measure money from the first week

    We set the baseline before anything is built and report value banked in the accounts. Usage and adoption are tracked, because they come first, but they are not the result we report to leadership. Every initiative has a business owner who signs off its numbers.

  7. 07

    We change the process first

    AI is a means. We redesign the work with the people who do it, with the tool in the middle of the new process. The business side usually has more to change than the technology side, and we say so at the start.

    We are also direct about limits. If AI will not move a number, or the data will not support it, we say so early and suggest what will.

  8. 08

    We hand the capability over

    Client staff join the team in the first week. Templates, playbooks, scoring models and dashboards belong to the client. By the second sprint, the client team runs most of the work and we advise. We also help hire the permanent people who will own it.

How the services connect

A typical relationship with an investor follows the life of an investment. During diligence we assess AI risk and upside. After closing, that view becomes a value-creation plan. The plan is delivered in sprints. A fractional lead runs the programme to agreed outcomes until a permanent leader takes it over. Our venture and research work keeps our methods current, and it feeds the next diligence.

  1. 01Diligence
  2. 02Value-creation plan
  3. 03Sprints
  4. 04Fractional lead
  5. 05Venture and research

Research feeds the next diligence.

If you want to see how a first sprint would look for your business, write to us.